All rights reserved. Unless otherwise specified, or required in the context of its implementation, no part of this publication may be reproduced or utilized otherwise in any form or by any means, electronic or mechanical, including photocopying, or posting on the internet or an intranet, without prior written permission. Permission can be requested from either ISO at the address below or ISO’s member body in the country of the requester. 著作權所有。除非另有規定,或在實施過程中需要,未經事先書面許可,不得以任何形式或任何方式(電子或機械,包括影印、在網際網路或內部網路發布)複製或利用本出版品的任何部分。可向以下地址的 ISO 或請求者所在國家/地區的 ISO 會員機構申請許可。
ISO copyright office ISO 著作權辦公室
CP 401 - Ch. de Blandonnet 8
CH-1214 Vernier, Geneva
Phone: +41 227490111 電話:+41 227490111
Fax: +41 227490947 傳真:+41 227490947
Email: copyright@iso.org 電子郵件:copyright@iso.org
Website: www.iso.org 網站:www.iso.org
Published in Switzerland 於瑞士出版
Contents 目錄
Page 頁次
Foreword … vi 前言 … vi
Introduction … vii 引言 … vii
1 Scope … 1 1 範圍 … 1
2 Normative references … 1 2 規範性引用文件 … 1
3 Terms and definitions … 2 3 術語與定義 … 2
4 Abbreviated terms … 7 4 縮寫詞 … 7
5 Conformance … 8 5 符合性 … 8
6 Modality specific information … 9 6 模態特定資訊 … 9
7 Abstract data elements … 9 7 抽象資料元素 … 9
7.1 Overview … 9 7.1 概述 … 9
7.1.1 Content and notation … 9 7.1.1 內容與標記法 … 9
7.1.2 Structure overview … 10 7.1.2 結構概述 … 10
7.1.3 Data conventions … 11 7.1.3 資料慣例 … 11
7.2 Face image data block … 11 7.2 臉部影像資料區塊 … 11
7.3 Version block … 11 7.3 版本區塊 … 11
7.4 Representation block … 12 7.4 表徵區塊 … 12
7.5 Representation identifier … 12 7.5 表徵識別碼 … 12
7.6 Capture date/time block … 12 7.6 擷取日期/時間區塊 … 12
7.7 Quality blocks … 12 7.7 品質區塊 … 12
7.8 PAD data block … 12 7.8 PAD 資料區塊 … 12
7.9 Session identifier … 12 7.9 工作階段識別碼 … 12
7.10 Derived from … 13 7.10 衍生自 … 13
7.11 Capture device block … 13 7.11 擷取裝置區塊 … 13
7.12 Model identifier block … 13 7.12 型號識別碼區塊 … 13
7.13 Certification identifier blocks … 13 7.13 認證識別碼區塊 … 13
7.14 Identity metadata block … 13 7.14 身分中繼資料區塊 … 13
7.15 Gender … 14 7.15 性別 … 14
7.16 Eye colour … 14 7.16 眼睛顏色 … 14
7.17 Hair colour … 15 7.17 頭髮顏色 … 15
7.18 Subject height … 15 7.18 主體身高 … 15
7.19 Properties block … 16 7.19 屬性區塊 … 16
7.20 Expression block … 16 7.20 表情區塊 … 16
7.21 Pose angle block … 17 7.21 姿態角度區塊 … 17
7.22 Angle data block … 18 7.22 角度資料區塊 … 18
7.23 Angle value … 18 7.23 角度值 … 18
7.24 Angle uncertainty … 18 7.24 角度不確定性 … 18
7.25 Landmark block … 19 7.25 地標區塊 … 19
7.26 Landmark kind … 19 7.26 地標類型 … 19
7.27 MPEG4 feature point … 19 7.27 MPEG4 特徵點 … 19
7.28 Anthropometric landmark … 22 7.28 人體測量學地標 … 22
7.29 Landmark coordinates block … 25 7.29 地標座標區塊 … 25
7.30 Image representation block … 26 7.30 影像表示區塊 … 26
7.31 2D image representation block … 26 7.31 2D 影像表示區塊 … 26
7.32 2D representation data … 26 7.32 2D 表示資料 … 26
7.33 2D capture device block … 26 7.33 2D 擷取裝置區塊 … 26
7.34 2D capture device spectral block … 26 7.34 2D 擷取裝置光譜區塊 … 26
7.35 2D capture device technology identifier … 27 7.35 2D 擷取裝置技術識別碼 … 27
7.36 2D image information block … 27 7.36 2D 影像資訊區塊 … 27
7.37 2D face image kind … 27 7.37 2D 臉部影像類型 … 27
7.38 Post acquisition processing block … 28 7.38 後期擷取處理區塊 … 28
7.39 Lossy transformation attempts … 28 7.39 有損轉換嘗試 … 28
7.40 Image data format … 29 7.40 影像資料格式 … 29
7.41 Camera to subject distance … 31 7.41 相機至主體距離 … 31
7.42 Sensor diagonal … 31 7.42 感測器對角線 … 31
7.43 Lens focal length … 32 7.43 鏡頭焦距 … 32
7.44 Image size block … 32 7.44 影像尺寸區塊 … 32
7.45 Width … 32 7.45 寬度 … 32
7.46 Height … 32 7.46 高度 … 32
7.47 Image face measurements block … 32 7.47 影像臉部測量區塊 … 32
7.48 Image head width … 33 7.48 影像頭部寬度 … 33
7.49 Image inter-eye distance … 33 7.49 影像雙眼間距 … 33
7.50 Image eye-to-mouth distance … 34 7.50 影像眼到嘴距離 … 34
7.51 Image head length … 34 7.51 影像頭部長度 … 34
7.52 Image colour space … 35 7.52 影像色彩空間 … 35
7.53 Reference colour mapping block … 35 7.53 參考色彩映射區塊 … 35
7.54 Reference colour schema … 35 7.54 參考色彩架構 … 35
7.55 Reference colour definition and value block … 35 7.55 參考色彩定義與數值區塊 … 35
7.56 3D shape representation block … 35 7.56 3D 形狀表示區塊 … 35
7.57 3D representation data … 36 7.57 3D 表示資料 … 36
7.58 3D capture device block … 36 7.58 3D 擷取裝置區塊 … 36
7.59 3D modus … 36 7.59 3D 模式 … 36
7.60 3D capture device technology identifier … 36 7.60 3D 擷取裝置技術識別碼 … 36
7.61 3D image information block … 37 7.61 3D 影像資訊區塊 … 37
7.62 3D representation kind block … 37 7.62 3D 表示類型區塊 … 37
7.63 3D vertex block … 37 7.63 3D 頂點區塊 … 37
7.64 3D vertex information block … 37 7.64 3D 頂點資訊區塊 … 37
7.65 3D vertex coordinate block … 37 7.65 3D 頂點座標區塊 … 37
7.66 3D vertex identifier … 38 7.66 3D 頂點識別碼 … 38
7.67 3D vertex normals block … 38 7.67 3D 頂點法線區塊 … 38
7.68 3D vertex textures block … 38 7.68 3D 頂點紋理區塊 … 38
7.69 3D error map … 38 7.69 3D 錯誤映射 … 38
7.70 3D vertex triangle data block … 39 7.70 3D 頂點三角形資料區塊 … 39
7.71 3D coordinate system … 39 7.71 3D 座標系統 … 39
7.72 3D Cartesian coordinate system … 39 7.72 3D 笛卡兒座標系統 … 39
7.73 3D Cartesian scales and offsets block … 40 7.73 3D 笛卡爾比例與偏移區塊 … 40
7.74 3D face image kind … 41 7.74 3D 臉部影像類型 … 41
7.75 3D physical face measurements block … 41 7.75 3D 實體臉部測量區塊 … 41
7.76 3D physical head width … 41 7.76 3D 實體頭部寬度 … 41
7.77 3D physical inter-eye distance … 42 7.77 3D 實體雙眼距離 … 42
7.78 3D physical eye-to-mouth distance … 42 7.78 3D 實體眼到嘴距離 … 42
7.79 3D physical head length … 42 7.79 3D 實體頭部長度 … 42
7.80 3D textured image resolution block … 42 7.80 3D 紋理影像解析度區塊 … 42
7.81 3D MM shape [X/Y/Z] resolution … 42 7.81 3D MM 形狀 [X/Y/Z] 解析度 … 42
7.82 3D MM texture resolution … 42 7.82 3D MM 紋理解析度 … 42
7.83 3D texture acquisition period … 42 7.83 3D 紋理擷取週期 … 42
7.84 3D face area scanned block … 43 7.84 3D 臉部區域掃描區塊 … 43
7.85 3D texture map block … 43 7.85 3D 紋理貼圖區塊 … 43
7.86 3D texture capture device spectral block … 43 7.86 3D 紋理擷取裝置光譜區塊 … 43
7.87 3D texture standard illuminant … 44 7.87 3D 紋理標準光源 … 44
7.88 3D texture map data … 44 7.88 3D 紋理貼圖資料 … 44
8 Encoding … 44 8 編碼 … 44
8.1 Overview … 44 8.1 概述 … 44
8.2 Tagged binary encoding … 45 8.2 標記二進位編碼 … 45
8.3 XML encoding … 46 8.3 XML 編碼 … 46
9 Registered BDB format identifiers … 47 9 已註冊的 BDB 格式識別碼 … 47
Annex A (normative) Formal specifications … 48 附錄 A (規範性) 正式規範 … 48
Annex B (informative) Encoding examples … 80 附錄 B (參考性) 編碼範例 … 80
Annex C (normative) Conformance testing methodology … 88 附錄 C (規範性) 一致性測試方法 … 88
Annex D (normative) Application profiles … 101 附錄 D (規範性) 應用設定檔 … 101
Annex E (informative) Additional technical considerations … 154 附錄 E (參考性) 其他技術考量 … 154
Bibliography … 184 參考書目 … 184
ISO/IEC 39794-5:2019(E)
Foreword 前言
ISO (the International Organization for Standardization) and IEC (the International Electrotechnical Commission) form the specialized system for worldwide standardization. National bodies that are members of ISO or IEC participate in the development of International Standards through technical committees established by the respective organization to deal with particular fields of technical activity. ISO and IEC technical committees collaborate in fields of mutual interest. Other international organizations, governmental and non-governmental, in liaison with ISO and IEC, also take part in the work. ISO(國際標準化組織)和 IEC(國際電工委員會)共同構成全球標準化的專門體系。ISO 或 IEC 的成員國機構透過各自組織為處理特定技術活動領域而設立的技術委員會,參與國際標準的制定。ISO 和 IEC 的技術委員會在共同感興趣的領域進行合作。其他與 ISO 和 IEC 保持聯繫的國際組織,無論是政府或非政府性質,也參與這項工作。
The procedures used to develop this document and those intended for its further maintenance are described in the ISO/IEC Directives, Part 1. In particular, the different approval criteria needed for the different types of document should be noted. This document was drafted in accordance with the editorial rules of the ISO/IEC Directives, Part 2 (see www.iso.org/directives). 本文件制定所使用的程序及其後續維護的預期程序,均詳述於 ISO/IEC 導則第 1 部分。特別應注意的是,不同類型文件所需的核准標準亦有所不同。本文件係依據 ISO/IEC 導則第 2 部分的編輯規則起草(詳見 www.iso.org/directives)。
Attention is drawn to the possibility that some of the elements of this document may be the subject of patent rights. ISO and IEC shall not be held responsible for identifying any or all such patent rights. Details of any patent rights identified during the development of the document will be in the Introduction and/or on the ISO list of patent declarations received (see www.iso.org/patents) or the IEC list of patent declarations received (see http://patents.iec.ch). 請注意,本文件的某些內容可能涉及專利權。ISO 和 IEC 對於識別任何或所有此類專利權不負任何責任。在文件制定過程中識別出的任何專利權詳細資訊,將載於引言和/或 ISO 已收到專利聲明清單(詳見 www.iso.org/patents)或 IEC 已收到專利聲明清單(詳見 http://patents.iec.ch)。
Any trade name used in this document is information given for the convenience of users and does not constitute an endorsement. 本文件所使用的任何商標名稱僅為方便使用者而提供,不構成任何形式的認可。
For an explanation of the voluntary nature of standards, the meaning of ISO specific terms and expressions related to conformity assessment, as well as information about ISO’s adherence to the World Trade Organization (WTO) principles in the Technical Barriers to Trade (TBT) see www.iso.org/ iso/foreword.html. 如需瞭解標準的自願性質、ISO 特定術語和與符合性評估相關表達的含義,以及 ISO 遵守世界貿易組織(WTO)在技術性貿易障礙(TBT)方面原則的資訊,請參閱 www.iso.org/iso/foreword.html。
This document was prepared by Joint Technical Committee ISO/IEC JTC 1, Information technology, Subcommittee SC 37, Biometrics. 本文件由 ISO/IEC JTC 1 資訊技術聯合技術委員會 SC 37 生物辨識小組委員會準備。
A list of all parts in the ISO/IEC 39794 series can be found on the ISO website. ISO/IEC 39794 系列的所有部分列表可在 ISO 網站上找到。
Any feedback or questions on this document should be directed to the user’s national standards body. A complete listing of these bodies can be found at www.iso.org/members.html. 任何關於本文件的回饋或問題,應向使用者所屬的國家標準機構提出。這些機構的完整列表可在 www.iso.org/members.html 找到。
The purchase of this ISO/IEC document carries a copyright licence for the purchaser to use ISO/IEC copyright in the schemas in the annexes to this document for the purpose of developing, implementing, installing and using software based on those schemas, subject to ISO/IEC licensing conditions set out in the schemas. 購買本 ISO/IEC 文件,即表示購買者獲得著作權許可,可將本文件附錄中架構的 ISO/IEC 著作權用於開發、實施、安裝及使用基於這些架構的軟體,但須遵守架構中規定的 ISO/IEC 授權條件。
Introduction 前言
Face images have been used for many decades to verify the identity of individuals. In recent years, digital face images have been used in many applications including human examination as well as computer-automated face recognition. Photographic formats are standardized, e.g., for passports and driver licences. There is also a need for a standard data format for digital face images to enable interoperability. A prominent case where such interoperability is essential is the electronic passport system, where face images are stored for several purposes including Automated Border Control. 臉部影像已用於驗證個人身份數十年。近年來,數位臉部影像已應用於許多領域,包括人工檢查以及電腦自動化臉部辨識。攝影格式已標準化,例如用於護照和駕照。此外,數位臉部影像也需要標準資料格式以實現互通性。電子護照系統是此類互通性至關重要的突出案例,其中臉部影像儲存用於多種目的,包括自動邊境管制。
Biometric data interchange formats enable the interoperability of different biometric systems. The first generation of biometric data interchange formats was published between 2005 and 2007 in the first edition of the ISO/IEC 19794 series. From 2011 onwards, the second generation of biometric data interchange formats was published in the second edition of the established parts and the first edition of some new parts of the ISO/IEC 19794 series. In the second generation of biometric data interchange formats, new useful data elements such as data elements related to biometric sample quality were added, the header data structures were harmonized across all parts of the ISO/IEC 19794 series, and XML encoding was been added in addition to the binary encoding. 生物特徵資料交換格式可實現不同生物辨識系統之間的互通性。第一代生物特徵資料交換格式於 2005 年至 2007 年間在 ISO/IEC 19794 系列的第一版中發布。從 2011 年起,第二代生物特徵資料交換格式在 ISO/IEC 19794 系列已確立部分的第二版和一些新部分的第一版中發布。在第二代生物特徵資料交換格式中,新增了有用的資料元素,例如與生物特徵樣本品質相關的資料元素,ISO/IEC 19794 系列所有部分的標頭資料結構也進行了協調,並且除了二進位編碼外,還增加了 XML 編碼。
In anticipation of the need for additional data elements, and in order to avoid future compatibility issues, the ISO/IEC 39794 series provides standard biometric data interchange formats capable of being extended in a defined way. Extensible specifications in ASN. 1 (Abstract Syntax Notation One) and the distinguished encoding rules (DER) of ASN. 1 form the basis for encoding biometric data in binary tag-length-value formats. XSDs (XML schema definitions) form the basis for encoding biometric data in XML (eXtensible Markup Language). 為預期對額外資料元素的需求,並避免未來的相容性問題,ISO/IEC 39794 系列提供了標準的生物特徵資料交換格式,能夠以定義的方式進行擴展。ASN.1(抽象語法標記法一)中的可擴展規範和 ASN.1 的區別編碼規則(DER)構成了以二進位標籤-長度-值格式編碼生物特徵資料的基礎。XSD(XML 綱要定義)構成了以 XML(可擴展標記語言)編碼生物特徵資料的基礎。
This third generation of face image data interchange formats complements ISO/IEC 19794-5:2005 and ISO/IEC 19794-5:2011. The first generation of biometric data interchange formats, which has been adopted, e.g., by ICAO for the biometric data stored in Machine Readable Travel Documents, is expected to be retained in the standards catalogue as long as needed. 第三代臉部影像資料交換格式補充了 ISO/IEC 19794-5:2005 和 ISO/IEC 19794-5:2011。第一代生物特徵資料交換格式已被國際民航組織(ICAO)等採用,用於儲存在機器可讀旅行證件中的生物特徵資料,預計將在標準目錄中保留,直至不再需要為止。
This document is intended to provide a generic face image data format for face recognition applications requiring exchange of face image data. Typical applications are: 本文件旨在為需要交換臉部影像資料的臉部辨識應用程式提供通用的臉部影像資料格式。典型的應用程式包括:
automated face biometric verification (one-to-one comparison) and identification (one-to-many comparison), and 自動化臉部生物特徵驗證(一對一比對)與辨識(一對多比對),以及
human verification of a biometric claim by comparison of data subjects against face images, including examination of face images with sufficient detail. 透過比對資料主體與臉部影像,進行生物特徵聲明的真人驗證,包括檢查具有足夠細節的臉部影像。
In addition to the data format, this document specifies application specific profiles including scene constraints, photographic properties and digital image attributes like image spatial sampling rate, image size, etc. These application profiles are contained in Annex D. 除了資料格式,本文件亦規定了應用程式專屬設定檔,包括場景限制、攝影屬性及數位影像屬性,例如影像空間取樣率、影像大小等。這些應用程式設定檔包含於附錄 D 中。
The structure of the data format in this document is not compatible with the previous generations. However, this new revision addresses, for the first time, a mechanism to maintain future extensions in a backwards- and forwards-compatible manner. This will mean that a parser is able to read data records and understand data items that are formatted according to versions of the standard that are older, the same or newer than the parser is developed to. All newer data items will not disrupt the parsing process and can be ignored. Newer versions of this document will at least include the mandatory data items of the previous standards. 本文件中的資料格式結構與前幾代不相容。然而,本次新修訂首次提出了一種機制,以向後和向前相容的方式維護未來的擴充功能。這表示解析器能夠讀取資料記錄並理解根據比解析器開發時更舊、相同或更新的標準版本所格式化的資料項目。所有較新的資料項目都不會中斷解析過程,並且可以被忽略。本文件的較新版本將至少包含先前標準的強制性資料項目。
The 3D encoding types 3D point map and range image are not supported by this edition of this document. 本文件此版本不支援 3D 編碼類型 3D 點圖和範圍影像。
Information technology - Extensible biometric data interchange formats - 資訊科技 — 可擴充生物特徵資料交換格式 —
Part 5:
Face image data 第 5 部分:臉部影像資料
1 Scope 1 範圍
This document specifies: 本文件規定:
generic extensible data interchange formats for the representation of face image data: A tagged binary data format based on an extensible specification in ASN. 1 and a textual data format based on an XML schema definition that are both capable of holding the same information; 用於表示臉部影像資料的通用可擴展資料交換格式:一種基於 ASN.1 中可擴展規範的標記二進位資料格式,以及一種基於 XML 綱要定義的文字資料格式,兩者都能夠包含相同的資訊;
examples of data record contents; 資料記錄內容範例;
application specific requirements, recommendations, and best practices in data acquisition; and 資料擷取中的應用特定要求、建議和最佳實踐;以及
conformance test assertions and conformance test procedures applicable to this document. 符合性測試斷言與適用於本文件的符合性測試程序。
2 Normative references 2 規範性引用文件
The following documents are referred to in the text in such a way that some or all of their content constitutes requirements of this document. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies. 下列文件內容構成本文件之部分或全部要求。對於註明日期之引用文件,僅適用所引用之版本。對於未註明日期之引用文件,則適用所引用文件之最新版本(包含任何修訂)。
ISO/IEC 2382-37, Information technology - Vocabulary - Part 37: Biometrics ISO/IEC 2382-37,資訊科技 — 詞彙 — 第 37 部分:生物辨識
ISO/IEC 8824-1, Information technology - Abstract Syntax Notation One (ASN.1): Specification of basic notation - Part 1 ISO/IEC 8824-1,資訊科技 — 抽象語法標記法一 (ASN.1):基本標記法規範 — 第 1 部分
ITU-T Rec. T. 81 | ISO/IEC 10918-1, Information technology - Digital compression and coding of continuous-tone still images - Part 1: Requirements and guidelines ITU-T 建議書 T.81 | ISO/IEC 10918-1,資訊科技 — 連續色調靜態影像的數位壓縮與編碼 — 第 1 部分:要求與指南
ISO 11664-2:2007, Colorimetry - Part 2: CIE standard illuminants ISO 11664-2:2007,色彩測量 — 第 2 部分:CIE 標準光源
ISO/IEC 14496-2:2004, Information technology - Coding of audio-visual objects - Part 2: Visual ISO/IEC 14496-2:2004,資訊科技—影音物件編碼—第 2 部分:視覺
ITU-T Rec. T. 800 | ISO/IEC 15444-1, Information technology - JPEG 2000 image coding system - Part 1: Core coding system ITU-T Rec. T. 800 | ISO/IEC 15444-1,資訊科技—JPEG 2000 影像編碼系統—第 1 部分:核心編碼系統
ISO/IEC 15948, Information technology - Computer graphics and image processing - Portable Network Graphics (PNG): Functional specification ISO/IEC 15948,資訊科技—電腦繪圖與影像處理—可攜式網路圖形(PNG):功能規範
ISO/IEC 39794-1, Information technology - Extensible biometric data interchange formats - Part 1: Framework ISO/IEC 39794-1,資訊科技 — 可擴展生物辨識資料交換格式 — 第 1 部分:框架
W3C Recommendation, XML Schema Part 2: Datatypes (Second Edition), 28 October 2004, http://www .w3.org/TR/xmlschema-2/ W3C 建議書,XML 綱要第 2 部分:資料型別 (第二版),2004 年 10 月 28 日,http://www .w3.org/TR/xmlschema-2/
3 Terms and definitions 3 術語與定義
For the purposes of this document, the terms and definitions given in ISO/IEC 39794-1, ISO/IEC 2382-37 and the following apply. 本文件適用 ISO/IEC 39794-1、ISO/IEC 2382-37 中給出的術語和定義,以及以下內容。
ISO and IEC maintain terminological databases for use in standardization at the following addresses: ISO 和 IEC 在以下網址維護用於標準化的術語資料庫:
ISO Online browsing platform: available at https://www.iso.org/obp ISO 線上瀏覽平台:可於 https://www.iso.org/obp 取得
3.1
1:1 application case 1:1 應用案例
biometric verification 生物識別驗證
Note 1 to entry: Biometric verification is defined in ISO/IEC 2382-37 as a process of confirming a biometric claim through biometric comparison. 條目註 1:生物識別驗證在 ISO/IEC 2382-37 中定義為透過生物識別比對來確認生物識別聲明的過程。
3.2
1:N application case 1:N 應用案例
biometric identification 生物識別
Note 1 to entry: Biometric identification is defined in ISO/IEC 2382-37 as a process of searching against a biometric enrolment database to find and return the biometric reference identifier(s) attributable to a single individual. 註 1:生物識別在 ISO/IEC 2382-37 中定義為一種針對生物識別註冊資料庫進行搜尋的過程,以尋找並傳回可歸因於單一受測者的生物識別參考識別碼。
3.3
2D face image 2D 臉部影像
two-dimensional face representation that encodes the luminance and/or colour texture of the face of a capture subject in a given lighting environment 在特定照明環境下,編碼擷取對象臉部亮度及/或色彩紋理的二維臉部表示
3.4
3D face image 3D 臉部影像
three-dimensional face representation that encodes a surface in a 3D space 在 3D 空間中編碼表面之三維臉部表示
3.5
3D vertex 3D 頂點
representation using 3D vertices and triangles between these points for coding of a 3D surface 使用 3D 頂點及這些點之間之三角形來編碼 3D 表面之表示
3.6
RGB
colour space designed to encompass most of the colours achievable on CMYK colour printers, but by using red, green and blue primary colours on a device such as a computer display 一種色彩空間,旨在涵蓋 CMYK 彩色印表機可實現的大部分顏色,但透過在電腦顯示器等裝置上使用紅、綠、藍三原色
3.7
anthropometric landmark 人體測量學地標
landmark on the face used for identification and classification of humans 用於人類識別和分類的臉部地標
3.8
landmark code 地標代碼
<anthropometric> two-part code that uniquely defines an anthropometric landmark <anthropometric> 唯一定義人體測量學地標的兩部分代碼
3.9
camera to subject distance 相機到受測者距離
CSD
distance between the eyes plane of a capture subject and the and the sensor/image plane of the camera 擷取對象的眼睛平面與相機感測器/影像平面之間的距離
3.10
Cartesian coordinate system 笛卡兒座標系統
3D orthogonal coordinate system 3D 正交座標系統
3.11
chin 下巴
central forward portion of the lower jaw 下顎中央靠前部分
3.12
CIE standard illuminant D65 CIE 標準光源 D65
commonly used standard illuminant defined by the International Commission on Illumination (CIE) that is intended to represent average daylight and has a correlated colour temperature of approximately 6500 K 國際照明委員會 (CIE) 定義的常用標準光源,旨在代表平均日光,其相關色溫約為 6500 K
Note 1 to entry: CIE standard illuminant D65 is specified in ISO 11664-2. 註 1:CIE 標準光源 D65 詳見 ISO 11664-2。
3.13
colour image 彩色影像
continuous tone image ( 3.16 _\underline{3.16} ) that has more than one channel, each of which is coded with one or multiple bits 具有一個以上通道的連續色調影像 ( 3.16 _\underline{3.16} ),每個通道都以一個或多個位元編碼
3.14
colour space 色彩空間
way of representing colours of pixels in an image 表示影像中像素顏色的方式
EXAMPLE RGB and YUV colour spaces are typically used in this document. 本文件通常使用 RGB 和 YUV 色彩空間。
3.15
common biometric exchange formats framework 通用生物識別交換格式框架
CBEFF
data format specifically for exchanging biometric data that provides for the encompassing of any biometric type into a standard format 專為交換生物識別資料而設計的資料格式,可將任何生物識別類型納入標準格式
3.16
continuous tone image 連續色調影像
image whose channels have more than one bit per pixel 每個像素的通道具有多於一個位元的影像
3.17
crop factor 裁切係數
ratio of the diagonal of the full frame camera (43,3mm)(43,3 \mathrm{~mm}) to that of a selected camera’s image sensor 全片幅相機 (43,3mm)(43,3 \mathrm{~mm}) 的對角線與所選相機影像感測器對角線之比
Note 1 to entry: The determination of an appropriate focal length lens for a field of view equivalent to a full frame camera can be made by considering the crop factor. 條目註 1:可透過考量裁切係數,來決定等同於全片幅相機視野的適當焦距鏡頭。
3.18
crown 冠部
top of the head ignoring any hair 頭頂(不含任何頭髮)
3.19
dots per inch 每英吋點數
DPI
individual printed dots in a line or column within a span of 25,4mm(1inch)25,4 \mathrm{~mm}(1 \mathrm{inch}) 在 25,4mm(1inch)25,4 \mathrm{~mm}(1 \mathrm{inch}) 的範圍內,一行或一列中個別列印的點
3.20
exposure value 曝光值
EV 曝光值
number that represents a combination of a camera’s shutter speed and f-number, such that all combinations that yield the same exposure have the same description value 代表相機快門速度和光圈值組合的數字,所有產生相同曝光的組合都具有相同的描述值
ISO/IEC 39794-5:2019(E)
3.21
eye centre 眼睛中心
centre of the line connecting the inner and the outer corner of the eye 連接眼睛內角和外角的線段中心
Note 1 to entry: The eye centres are the feature points 12.1 and 12.2 as defined in ISO/IEC 14496-2:2004, Annex C. 註記 1:眼睛中心是 ISO/IEC 14496-2:2004 附錄 C 中定義的特徵點 12.1 和 12.2。
Note 2 to entry: The inner and the outer corner of the eye are defined by ISO/IEC 14496-2-2:2004: feature points 3.12 and 3.8 for the right eye, and 3.11 and 3.7 for the left eye. 條目 2 的註釋:眼睛的內眼角和外眼角由 ISO/IEC 14496-2-2:2004 定義:右眼為特徵點 3.12 和 3.8,左眼為 3.11 和 3.7。
3.22
eye-to-mouth distance 眼到嘴距離
EMD
distance between the face centre and the mouth midpoint 臉部中心與嘴部中點之間的距離
Note 1 to entry: The mouth midpoint is the feature point 2.3 as defined in ISO/IEC 14496-2:2004, Annex C. 註記 1:嘴部中點為 ISO/IEC 14496-2:2004 附錄 C 中定義的特徵點 2.3。
3.23
eye visibility zone 眼睛可視區域
EVZ
zone covering a rectangle having a margin to any part of the visible eyeball 涵蓋矩形區域,其邊緣與可見眼球的任何部分皆有間距
Note 1 to entry: The margin is defined in D.1.4.3.3. 註記 1 至條目:邊界定義於 D.1.4.3.3。
3.24
face centre 臉部中心
M
midpoint of the line connecting the two eye centres 連接兩眼中心的線段中點
3.25
face image kind 臉部影像類型
category of face images ( 3.27 _\underline{3.27} ) that satisfy specific requirements 符合特定要求之臉部影像類別 ( 3.27 _\underline{3.27} )
Note 1 to entry: Application specific requirements are specified in one of the application profiles in Annex D. 註 1:應用程式特定要求詳載於附錄 D 的應用程式設定檔之一。
3.26
facial animation parameter 臉部動畫參數
FAP
standard for the virtual representation, which includes visual speech intelligibility, mood and gesture by using feature points 用於虛擬呈現的標準,包括透過使用特徵點來呈現視覺語音清晰度、情緒和手勢
Note 1 to entry: Visual representation as specified in ISO/IEC 14496-1 and ISO/IEC 14496-2. 註記 1:視覺呈現如 ISO/IEC 14496-1 和 ISO/IEC 14496-2 所述。
3.27
face image 臉部影像
electronic image-based representation of the face of a capture subject 擷取對象臉部的電子影像式呈現
3.28
face portrait 臉部肖像
visual representation of the capture subject, which includes the full-frontal part of the head, including hair in most cases, as well as neck and possibly top of shoulders 捕捉對象的視覺呈現,包括頭部的正前方部分,在大多數情況下包括頭髮,以及頸部和可能的部分肩膀
3.29
face texture 臉部紋理
2D sampling face representation that encodes one or a combination of several spectral spatial modulations received by 3D imaging systems of a face in a given lighting system having a 2D coordinate link to the face shape 2D 採樣臉部表示,其編碼在給定照明系統中,由 3D 成像系統接收到的一個或多個光譜空間調變的組合,該調變具有與臉部形狀相關的 2D 座標連結
3.30
feature point 特徵點
reference point in a face image as used by face recognition algorithms 臉部影像中供臉部辨識演算法使用的參考點
Note 1 to entry: Commonly referred to as a landmark, an example being the position of the eyes. 註 1:通常稱為地標,例如眼睛的位置。
3.31
fish eye 魚眼
type of distortion where central objects of the image erroneously appear closer than those at the edge 一種扭曲類型,其中影像的中心物體錯誤地顯得比邊緣物體更近
3.32
Frankfurt Horizon 法蘭克福水平線
standard plane for orientation of the head defined by a line passing through the right tragion (the front of the ear) and the lowest point of the right eye socket 一種用於頭部定向的標準平面,由一條通過右耳屏(耳朵前方)和右眼窩最低點的線定義
Note 1 to entry: The Frankfurt Horizon may be hard to define, as it is related to the ear position that may be covered by hair. 條目註 1:法蘭克福平面可能難以定義,因為它與可能被頭髮覆蓋的耳朵位置有關。
Note 2 to entry: The Frankfurt Horizon has been defined in the Frankfurt-am-Main (anthropological) Agreement of 1882. 條目註 2:法蘭克福水平線(Frankfurt Horizon)已於 1882 年的法蘭克福(人類學)協議中定義。
3.33
greyscale image 灰階影像
continuous tone image ( 3.16 _\underline{3.16} ) encoded with one luminance channel 以一個亮度通道編碼的連續色調影像( 3.16 _\underline{3.16} )
Note 1 to entry: If the luminance channel is coded with 8 bits, the greyscale image is also referred to as a monochrome or black and white image. 條目註 1:如果亮度通道以 8 位元編碼,則灰階影像也稱為單色或黑白影像。
3.34
horizontal deviation angle 水平偏差角
HD 高畫質
maximal allowed deviation from the horizontal of the imaginary line between the nose of a capture subject and the lens of the camera 擷取對象的鼻子與相機鏡頭之間假想線的水平最大允許偏差
3.35
human examination 人工檢視
process of human comparison of a face image with an individual or another face image through detailed examination of face characteristics and structures for the purposes of biometric verification or identification 透過詳細檢視臉部特徵和結構,將臉部影像與個人或另一臉部影像進行人工比對的過程,以達到生物識別驗證或辨識的目的。
3.36
human identification 人體識別
process of human searching through a list of face images to match against an input image(s) 人工搜尋臉部影像清單以比對輸入影像的過程。
Note 1 to entry: Also known as one-to-many (1:N) searching. 註 1:亦稱為一對多 (1:N) 搜尋。
Note 2 to entry: Identification can be performed by human (experts) as well, and human identification may consider more than biometric data. 條目 2 的註釋:識別也可以由人類(專家)執行,且人類識別可能考慮的不僅是生物特徵資料。
3.37
human verification 人體驗證
process of confirming a specific biometric claim by human comparison of a face image with an individual or another face image 透過人臉影像與個人或另一張人臉影像進行人工比對,以確認特定生物特徵聲明的過程
Note 1 to entry: Also known as one-to-one ( 1:11: 1 ) comparison. 條目 1 的註釋:亦稱為一對一( 1:11: 1 )比對。
Note 2 to entry: Verification can be performed by human (experts) as well, and human verification may consider more than biometric data. 條目 2 的註記:驗證亦可由人類(專家)執行,且人類驗證可能考量生物特徵資料以外的資訊。
3.38
implementation under test
IUT
Implementation of a technical system currently tested 3.38 測試中實作 IUT 目前正在測試的技術系統實作
3.39
inner region 內部區域
pixels of a face image carrying data of the central region of a face 臉部影像中承載臉部中央區域資料的像素
ISO/IEC 39794-5:2019(E)
3.40
inter-eye distance 眼間距
IED
length of the line connecting the eye centres of the left and right eye 連接左右眼中心點的線段長度
3.41
issuer 發行者
organization that issues Machine Readable Travel Documents (MRTDs) 發行機器可讀旅行證件(MRTD)的組織
3.42
lower camel-case notation 小駝峰式命名法
naming convention in which compound words are joined together without spaces, where the first letter of the entire word is lowercase, but the first letter of subsequent words is uppercase 一種命名慣例,其中複合詞不含空格地連接在一起,整個單字的第一個字母為小寫,但後續單字的第一個字母為大寫
3.43
magnification distortion 放大失真
image imperfection where the degree of magnification varies with the distance from the camera and the depth of the face 影像缺陷,其放大程度隨與攝影機的距離和臉部的深度而變化
3.44
modus 模式
manner in which a particular property is acquired 取得特定屬性的方式
3.45
near infrared 近紅外線
section of infrared band with wavelength from 780 nm to 3000 nm 波長介於 780 奈米至 3000 奈米之間的紅外線波段
3.46
outer region 外部區域
pixels of a face image outside of the inner region 臉部影像中位於內部區域外的像素
3.47
photo booth 快照亭
automated system for digitally capturing 2D images in either public or office environments 用於在公共或辦公環境中數位擷取二維影像的自動化系統
Note 1 to entry: A photo booth encloses the subject in a highly-controlled lighting environment and consists of a camera, lighting and peripheral devices such as printers. It has entrances on one or both sides with reflective curtains protecting against ambient light. 註記 1:照相亭將拍攝對象置於高度受控的照明環境中,並由相機、照明和印表機等周邊設備組成。它在一側或兩側設有入口,並配有反光窗簾以防止環境光線。
3.48
photo kiosk 相片快照亭
semi-automated system for digitally capturing 2D images in an office-environment 用於在辦公環境中數位擷取二維影像的半自動化系統
Note 1 to entry: A photo kiosk consists of camera and lighting and usually has a separate panel placed behind the subject to provide the required background but is otherwise open. 備註 1:照相亭由相機和照明組成,通常在拍攝對象後方設有獨立的面板,以提供所需的背景,但其他部分是開放的。
3.49
pixel 像素
picture element on a two-dimensional array that comprises an image 構成影像的二維陣列上的圖像元素
3.50
pixel per inch 每英吋像素
PPI
individual pixels in a line or column of a digital image within a span of 25,4mm(1inch)25,4 \mathrm{~mm}(1 \mathrm{inch}) 在 25,4mm(1inch)25,4 \mathrm{~mm}(1 \mathrm{inch}) 的範圍內,數位影像中某行或某列的個別像素
3.51
presentation attack 呈現攻擊
presentation of an artefact or human characteristic to the biometric capture subsystem in a fashion that could interfere with the intended policy of the biometric system 以可能干擾生物識別系統預期政策的方式,向生物識別擷取子系統呈現人工製品或人類特徵
3.52
presentation attack detection 呈現攻擊偵測
PAD
automated determination of a presentation attack 自動判斷呈現攻擊
3.53
radial distortion 徑向畸變
image imperfection where the degree of magnification varies with the distance from the optical axis 影像缺陷,其放大程度隨與光軸的距離而變化
3.54
red eye effect 紅眼效應
red glow from a subject’s eye caused by light from flash reflecting from blood vessels behind the retina 主體眼睛因閃光燈光線從視網膜後方血管反射而產生的紅光
3.55
subject 主體
individual who is to be displayed on the face portrait 臉部肖像上顯示的個人
Note 1 to entry: If the face portrait is part of a Machine Readable Travel Document (MRTD), this individual is intended to be the holder of the MRTD. 註 1:如果臉部肖像為機器可讀旅行證件 (MRTD) 的一部分,則此個人應為 MRTD 的持有人。
3.56
upper camel-case notation 大駝峰式命名法
naming convention in which compound words are joined together without spaces and the first letter of every word is uppercase 一種命名慣例,其中複合詞不含空格地連接在一起,且每個單字的第一個字母均為大寫
3.57
wavelength 波長
distance between repeating units of a wave pattern 波型重複單元之間的距離
Note 1 to entry: Commonly designated by the Greek letter lambda ( lambda\lambda ). 註記 1:通常以希臘字母 lambda ( lambda\lambda ) 表示。
3.58
white light 白光
apparently colourless light on human perception 人眼感知中近似無色的光線
EXAMPLE Ordinary daylight, standard lights as D50, D65, etc. 範例 一般日光、標準光源,例如 D50、D65 等。
Note 1 to entry: For many purposes it is assumed that white light contains all wavelengths of the visible spectrum at equal intensity based on human perception. Strong deviations from equal intensity usually lead to deviations in the perception of colours. 註記 1:就許多目的而言,假設白光包含可見光譜中所有波長,且基於人眼感知,其強度相等。強度若有顯著偏差,通常會導致顏色感知上的偏差。
4 Abbreviated terms 4 縮寫詞
For the purposes of this document, the abbreviated terms given in ISO/IEC 39794-1 and the following apply. 為本文件之目的,適用 ISO/IEC 39794-1 中給出的縮寫詞及以下內容。
ABC Automated Border Control ABC 自動化邊境查驗
CCD charge-coupled device CCD 電荷耦合元件
CMOS complementary metal-oxide-semiconductor CMOS 互補式金屬氧化物半導體
CSD camera to subject distance CSD 攝影機到受測者距離
DOVID diffractive optically variable image device DOVID 繞射光學可變影像裝置
DPI dots per inch DPI 每英吋點數
EMD eye-to-mouth distance EMD 眼到嘴距離
ISO/IEC 39794-5:2019(E)
EV
exposure value 曝光值
EVZ
eye visibility zone 眼睛可視區
FAP
facial animation parameter 臉部動畫參數
FH
Frankfurt Horizon 法蘭克福水平
HD
horizontal deviation angle 水平偏差角
ICAO
International Civil Aviation Organization 國際民用航空組織
IED
Inter-eye distance 瞳距
JPEG
image compression standard specified as ISO/IEC 10918; the JPEG baseline standard was published as ITU-T Rec. T. 81 | ISO/IEC 10918-1 影像壓縮標準,由 ISO/IEC 10918 指定;JPEG 基準標準發布為 ITU-T Rec. T. 81 | ISO/IEC 10918-1
JPEG2000
image compression standard specified as ISO/IEC 15444; the JPEG2000 baseline standard was published as ITU-T Rec. T. 800 | ISO/IEC 15444-1 影像壓縮標準,定義為 ISO/IEC 15444;JPEG2000 基本標準發布為 ITU-T Rec. T. 800 | ISO/IEC 15444-1
LDS
logical data structure as defined in ICAO Doc 9303 邏輯資料結構,定義於 ICAO Doc 9303
M
face centre 臉部中心
MP
intensity measurement pattern side length 強度量測圖案邊長
MRTD
machine readable travel document, the term also includes electronic MRTD's, electronic machine readable travel document using a contactless integrated circuit 機器可讀旅行文件,該術語亦包含電子式機器可讀旅行文件,使用非接觸式積體電路的電子式機器可讀旅行文件
MTF
modulation transfer function 調變轉換函數
MTF20
highest spatial frequency where the MTF is 20%20 \% or above MTF 為 20%20 \% 或以上之最高空間頻率
NIR
near infrared 近紅外線
PPCM
pixel per centimetre 每公分像素
PPI
pixel per inch 每英吋像素
PNG
portable network graphics format specified as ISO/IEC 15948 可攜式網路圖形格式,依據 ISO/IEC 15948 規範
RFID
radio-frequency identification 無線射頻辨識
RGB
red green blue colour representation 紅綠藍色彩表示
SFR
spatial frequency response 空間頻率回應
SNR
signal to noise ratio 訊號雜訊比
sRGB
standard RGB colour space created for use on monitors, printers and the Internet using the ITU-R BT. 709 primaries 為顯示器、印表機和網際網路使用而建立的標準 RGB 色彩空間,採用 ITU-R BT. 709 原色
EV exposure value
EVZ eye visibility zone
FAP facial animation parameter
FH Frankfurt Horizon
HD horizontal deviation angle
ICAO International Civil Aviation Organization
IED Inter-eye distance
JPEG image compression standard specified as ISO/IEC 10918; the JPEG baseline standard was published as ITU-T Rec. T. 81 | ISO/IEC 10918-1
JPEG2000 image compression standard specified as ISO/IEC 15444; the JPEG2000 baseline standard was published as ITU-T Rec. T. 800 | ISO/IEC 15444-1
LDS logical data structure as defined in ICAO Doc 9303
M face centre
MP intensity measurement pattern side length
MRTD machine readable travel document, the term also includes electronic MRTD's, electronic machine readable travel document using a contactless integrated circuit
MTF modulation transfer function
MTF20 highest spatial frequency where the MTF is 20% or above
NIR near infrared
PPCM pixel per centimetre
PPI pixel per inch
PNG portable network graphics format specified as ISO/IEC 15948
RFID radio-frequency identification
RGB red green blue colour representation
SFR spatial frequency response
SNR signal to noise ratio
sRGB standard RGB colour space created for use on monitors, printers and the Internet using the ITU-R BT. 709 primaries| EV | exposure value |
| :--- | :--- |
| EVZ | eye visibility zone |
| FAP | facial animation parameter |
| FH | Frankfurt Horizon |
| HD | horizontal deviation angle |
| ICAO | International Civil Aviation Organization |
| IED | Inter-eye distance |
| JPEG | image compression standard specified as ISO/IEC 10918; the JPEG baseline standard was published as ITU-T Rec. T. 81 \| ISO/IEC 10918-1 |
| JPEG2000 | image compression standard specified as ISO/IEC 15444; the JPEG2000 baseline standard was published as ITU-T Rec. T. 800 \| ISO/IEC 15444-1 |
| LDS | logical data structure as defined in ICAO Doc 9303 |
| M | face centre |
| MP | intensity measurement pattern side length |
| MRTD | machine readable travel document, the term also includes electronic MRTD's, electronic machine readable travel document using a contactless integrated circuit |
| MTF | modulation transfer function |
| MTF20 | highest spatial frequency where the MTF is $20 \%$ or above |
| NIR | near infrared |
| PPCM | pixel per centimetre |
| PPI | pixel per inch |
| PNG | portable network graphics format specified as ISO/IEC 15948 |
| RFID | radio-frequency identification |
| RGB | red green blue colour representation |
| SFR | spatial frequency response |
| SNR | signal to noise ratio |
| sRGB | standard RGB colour space created for use on monitors, printers and the Internet using the ITU-R BT. 709 primaries |
5 Conformance 5 符合性
A BDB conforms to this document if it satisfies all the relevant requirements related to: 如果 BDB 滿足所有相關要求,則符合本文件,這些要求與:
its data structure, data values and the relationships between its data elements as specified throughout Clauses 7,8 and Annex A, 其資料結構、資料值以及其資料元素之間的關係,如第 7、8 條和附錄 A 中所規定,
the relationship between its data values and the input biometric data from which the BDB was generated as specified throughout Clauses 7,8 , and Annex A, and 其資料值與產生 BDB 的輸入生物識別資料之間的關係,如第 7、8 條和附錄 A 中所規定,以及
application profile specific compliance specifications given in Annex C.4. 附錄 C.4 中給出的應用程式個人資料特定合規規範。
A system that produces biometric data records is conformant to this document if all biometric data records that it outputs conform to this document (as defined above) as claimed in the ICS associated with that system. A system does not need to be capable of producing biometric data records that cover all possible aspects of this document, but only those that are claimed to be supported by the system in the ICS. 如果一個系統所輸出的所有生物識別資料記錄都符合本文件(如上所述),並在與該系統相關的 ICS 中聲明,則該系統符合本文件。一個系統不需要能夠產生涵蓋本文件所有可能方面的生物識別資料記錄,而只需產生在 ICS 中聲明該系統支援的那些方面。
A system that uses BDBs is conformant to this document if it can read, and use for the purpose intended by that system, all BDBs that conform to this document (as defined above) as claimed in the ICS associated with that system. A system does not need to be capable of using BDBs that cover all possible aspects of this document, but only those that are claimed to be supported by the system in an ICS. 如果一個使用 BDB 的系統能夠讀取並用於該系統預期目的的所有符合本文件(如上所述)的 BDB,並在與該系統相關的 ICS 中聲明,則該系統符合本文件。一個系統不需要能夠使用涵蓋本文件所有可能方面的 BDB,而只需使用在 ICS 中聲明該系統支援的那些方面。
6 Modality specific information 6 模態特定資訊
The recorded image data shall appear to be the result of a capture process of a face. For the purpose of describing the position of each pixel within an image to be exchanged, a pair of reference axes shall be used. The origin of the axes, pixel location ( 0,0 ), shall be located at the upper left-hand corner of each image, which corresponds to the upper right-hand side of the forehead from the perspective of the capture subject. The xx-coordinate (horizontal) position shall increase positively from the origin to the right side of the image (i.e. left-hand forehead). The yy-coordinate (vertical) position shall increase positively from the origin to the bottom of the image. 所記錄的影像資料應呈現為人臉擷取過程的結果。為描述影像中每個像素的交換位置,應使用一對參考軸。軸的原點,像素位置 (0,0),應位於每個影像的左上角,這從擷取對象的角度來看,對應於額頭的右上側。 xx 座標(水平)位置應從原點向影像右側(即左側額頭)正向增加。 yy 座標(垂直)位置應從原點向影像底部正向增加。
7 Abstract data elements 7 抽象資料元素
7.1 Overview 7.1 概述
7.1.1 Content and notation 7.1.1 內容與標記法
This clause describes the contents of data elements defined in this document. These semantic descriptions are independent of the encoding of the data elements. 本條款描述了本文件中定義的資料元素的內容。這些語義描述與資料元素的編碼無關。
The presence of data elements is specified in Annex A. Certain data elements are optional. Such data elements need not be included in a BDB. An optional data element may be omitted altogether from the encoding. 資料元素的存在性在附錄 A 中指定。某些資料元素是可選的。此類資料元素無需包含在 BDB 中。可選資料元素可以完全從編碼中省略。
Application profiles as defined in Annex D may further restrict the presence of data elements. Such profiles may make optional elements mandatory, and they may exclude optional elements. 附錄 D 中定義的應用程式設定檔可以進一步限制資料元素的存在性。此類設定檔可以使可選元素成為強制性,並且可以排除可選元素。
In an ASN. 1 module, optional data elements are marked with the keyword OPTIONAL. When such an element is not present, then the tag, length and value octets of this data element are omitted from the tagged binary encoding. 在 ASN.1 模組中,可選資料元素會以關鍵字 OPTIONAL 標記。當此類元素不存在時,則此資料元素的標籤、長度及值位元組會從已標記的二進位編碼中省略。
A data element in an XML schema definition is optional if the value of its min0ccurs attribute is 0 . When such an element is not present, the opening and closing tags as well as the value of this data element are omitted from the XML encoding. XML 綱要定義中的資料元素,如果其 minOccurs 屬性值為 0,則為可選。當此類元素不存在時,XML 編碼中會省略其開頭和結尾標籤以及此資料元素的值。
If all child elements of a data element are optional, this data element shall be marked optional as well. 如果資料元素的所有子元素都是可選的,則此資料元素也應標記為可選。
Type names are in upper camel-case notation derived from subclause titles in this clause. Element names are in lower camel-case notation derived from these subclause titles. If the generic name starts with a number, then this component is set to the end of the base name. In the XSD, type names will end with the word “Type”. 類型名稱採用大駝峰式命名法,源自本條款中的子條款標題。元素名稱採用小駝峰式命名法,源自這些子條款標題。如果通用名稱以數字開頭,則此元件會設定在基本名稱的末尾。在 XSD 中,類型名稱將以「Type」一詞結尾。
EXAMPLE 1 The Image Colour Space element has the encoding name imageColourSpace and the type ImageColourSpace (in ASN.1) and ImageColourSpaceType (in XML). 範例 1 影像色彩空間元素具有編碼名稱 imageColourSpace 和類型 ImageColourSpace (在 ASN.1 中) 以及 ImageColourSpaceType (在 XML 中)。
EXAMPLE 2 An element value with the abstract name colour coded light has the value colourCodedLight. An element value with the abstract name 48 bit RGB has the encoding value rgb48Bit. 範例 2 抽象名稱為 colour coded light 的元素值為 colourCodedLight。抽象名稱為 48 bit RGB 的元素值為 rgb48Bit。
7.1.2 Structure overview 7.1.2 結構概述
The order of the abstract data elements in 7.2 and beyond is derived from traversing the tree in Figure 1 from left to right, depth first. A formal description of the structure is given in Annex A. 1 for ASN. 1 and in Annex A. 2 for the XML encoding of these abstract data elements. 7.2 及以後的抽象資料元素順序是透過從左到右、深度優先遍歷圖 1 中的樹狀結構而得出的。結構的正式描述在附錄 A.1 中針對 ASN.1 給出,在附錄 A.2 中針對這些抽象資料元素的 XML 編碼給出。
Key 鍵
elements which can be divided in sub-elements, not shown in this Figure 可分為子元素的元素,未顯示於本圖
Exclusive Or (XOR), one, and only one, option shall be chosen 互斥或(XOR),應選擇且僅選擇一個選項
denotes that this element is defined in Clause 7.n 表示此元素定義於條款 7.n
The Figure has been manually generated, its content is informative. The normative structure is given in A. 1 for ASN. 1 and A. 2 for XML. 本圖為手動生成,其內容僅供參考。規範性結構在 A.1 中用於 ASN.1,在 A.2 中用於 XML。
Figure 1 - Face image data block 圖 1 - 臉部影像資料區塊
7.1.3 Data conventions 7.1.3 資料慣例
For value measurement the following units are used: 數值測量使用以下單位:
physical measurement: millimetres; 實體測量:公釐;
image measurement: pixels; 影像測量:像素;
left/right: from perspective of the subject. 左/右:從主體的角度來看。
Unless otherwise specified, all other numeric values are unsigned integer quantities. 除非另有規定,所有其他數值均為無符號整數。
The conversion of a numeric value to integer is given by rounding down if the fractional portion is less than 0,5 and rounding up if the fractional value is greater than or equal to 0,5 . 數值轉換為整數的規則為:若小數部分小於 0.5,則向下取整;若小數部分大於或等於 0.5,則向上取整。
The absence of an optional element means that the encoder does not provide any statement about the value of the element. 若選用元素不存在,表示編碼器未提供該元素值的任何說明。
7.2 Face image data block 7.2 臉部影像資料區塊
Figure 2 - Example of embedding multiple representations in the same Face image data block 圖 2 - 在相同臉部影像資料區塊中嵌入多個表示法的範例
Abstract values: None 抽象值:無
Contents: Each BDB shall pertain to a single subject and shall contain one or more representations of a human face. Together with the Version block, each BDB can contain one or more geometric representations in Representation blocks. The record structure is depicted in Figure 2. 內容:每個 BDB 應與單一主體相關,並應包含一個或多個人臉的表示法。連同版本區塊,每個 BDB 可以在表示法區塊中包含一個或多個幾何表示法。記錄結構如圖 2 所示。
7.3 Version block 7.3 版本區塊
Abstract values: The abstract values for the Version block are defined in ISO/IEC 39794-1. 抽象值:版本區塊的抽象值定義於 ISO/IEC 39794-1。
Contents: The generation number of this document shall be 3. The year shall be the year of the publication of this document. 內容:本文件的產生編號應為 3。年份應為本文件發布的年份。
If a BDB contains representations encoded using different versions of an extensible biometric data interchange format, then the version number of the most recent version of the encoding versions shall be used. 如果 BDB 包含使用不同版本可擴展生物識別資料交換格式編碼的表示,則應使用編碼版本中最新版本的版本號。
7.4 Representation block 7.4 表徵區塊
Abstract values: None. 抽象值:無。
Contents: A Representation block consists of a unique Representation identifier characterizing this Representation block, an Image representation block, a Capture date/time block, Quality blocks, a PAD data block, a Session identifier, an identifier to define a relationship to another record, called Derived from, a Capture device block, a Identity metadata block describing discernible characteristics of the subject, and the Landmark blocks. The structure of this element is shown in Figure 1. 內容:表徵區塊包含一個獨特的表徵識別碼,用於識別此表徵區塊、一個影像表徵區塊、一個擷取日期/時間區塊、品質區塊、一個 PAD 資料區塊、一個工作階段識別碼、一個用於定義與另一筆記錄(稱為「衍生自」)關係的識別碼、一個擷取裝置區塊、一個描述受試者可辨識特徵的身分中繼資料區塊,以及地標區塊。此元素的結構如圖 1 所示。
Multiple face image representations of the same biometric data subject may be described in the same Face image data block. This is accomplished by including multiple Representation blocks. Face image representations containing 2D data may be combined with face image representations containing 3D data. 同一個生物識別資料受試者的多個臉部影像表徵,可以在同一個臉部影像資料區塊中描述。這可透過包含多個表徵區塊來實現。包含 2D 資料的臉部影像表徵可與包含 3D 資料的臉部影像表徵結合。
EXAMPLE The structure of a possible storage of Respresentaton blocks containing 2D and 3D data is illustrated in Figure 2. 範例 包含 2D 和 3D 資料的表徵區塊可能儲存結構如圖 2 所示。
7.5 Representation identifier 7.5 表徵識別碼
Abstract values: Integer. 抽象值:整數。
Contents: This element shall obtain a unique identifier for the Representation block. Each representation shall have its unique Representation identifier. 內容:此元素應取得表徵區塊的唯一識別碼。每個表徵都應有其唯一的表徵識別碼。
NOTE Unlike other parts of the ISO/IEC 39794 series, this document requires Representation identifiers to link processed data to its original source. 註:與 ISO/IEC 39794 系列的其他部分不同,本文件要求表徵識別碼將處理過的資料連結至其原始來源。
7.6 Capture date/time block 7.6 擷取日期/時間區塊
Abstract values: See Capture date/time block in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的擷取日期/時間區塊。
Contents: The Capture date/time block shall indicate when the capture of this representation started in Coordinated Universal Time (UTC). 內容:擷取日期/時間區塊應以世界協調時間(UTC)表示此呈現開始擷取的時間。
7.7 Quality blocks 7.7 品質區塊
Abstract values: See Quality blocks in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的品質區塊。
Contents: This element contains information on the biometric sample quality. 內容:此元素包含生物特徵樣本品質的資訊。
7.8 PAD data block 7.8 PAD 資料區塊
Abstract values: See PAD data block in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的 PAD 資料區塊。
Contents: This element shall convey the mechanism used in biometric presentation attack detection and the results of the presentation attack detection mechanism. 內容:此元素應傳達生物特徵呈現攻擊偵測中使用的機制,以及呈現攻擊偵測機制的結果。
7.9 Session identifier 7.9 工作階段識別碼
Abstract values: Integer. 抽象值:整數。
Contents: This element shall map the Representation block to the photo session where the face image was recorded. 內容:此元素應將表示區塊對應至記錄臉部影像的拍照工作階段。
7.10 Derived from 7.10 衍生自
Abstract values: Integer. 抽象值:整數。
Contents: This element shall denote interdependencies when multiple representations are stored in a Face image data block. This is of particular interest in the case where post-processing has been used but may be used in case of all other image types, too. The value shall be the Representation identifier number of the original representation. 內容:當多個表示儲存在臉部影像資料區塊中時,此元素應表示其相互依賴性。這在已使用後處理的情況下特別重要,但也可應用於所有其他影像類型。其值應為原始表示的表示識別碼。
To give an example for an application of this specification, assume that there are two Representations in the overall record. Their identifiers are 1 and 2. The first representation has been post-processed and resulted in the second representation. Then, the second representation shall have the Derived from element set to 1 . 為了舉例說明本規範的應用,假設整體記錄中有兩個表示。它們的識別碼分別是 1 和 2。第一個表示經過後處理後產生了第二個表示。那麼,第二個表示的「Derived from」元素應設定為 1。
7.11 Capture device block 7.11 擷取裝置區塊
Abstract values: See Capture device block in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的擷取裝置區塊。
Contents: The Capture device block contains the Model identifier block and the Certification identifier blocks. 內容:擷取裝置區塊包含型號識別碼區塊和認證識別碼區塊。
7.12 Model identifier block 7.12 模型識別碼區塊
Abstract values: See Model identifier block in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的模型識別碼區塊。
Contents: The Model identifier block shall identify the biometric organization that manufactures the product that created the BDB . It shall carry a CBEFF biometric organization identifier (see ISO/IEC 39794-1). Additionally, it shall identify the product type that created the BDB. It shall be assigned by the registered product manufacturer or other approved registration authority (see ISO/IEC 39794-1). 內容:模型識別碼區塊應識別製造建立 BDB 產品的生物識別組織。它應包含 CBEFF 生物識別組織識別碼(請參閱 ISO/IEC 39794-1)。此外,它應識別建立 BDB 的產品類型。它應由註冊產品製造商或其他經核准的註冊機構指定(請參閱 ISO/IEC 39794-1)。
7.13 Certification identifier blocks 7.13 認證識別碼區塊
Abstract values: See Certification identifier blocks in ISO/IEC 39794-1. 抽象值:請參閱 ISO/IEC 39794-1 中的認證識別碼區塊。
Contents: This document does not contain details of certification schemes. 內容:本文件不包含認證方案的詳細資訊。
NOTE Currently, no certification schemes are available for this document. 註:目前,本文件尚無可用的認證方案。
7.14 Identity metadata block 7.14 身份中繼資料區塊
Abstract values: For the structure see Figure 1. 抽象值:結構請參閱圖 1。
Contents: The Identity metadata block is intended to describe properties of the subject pictured in the image. The Identity metadata block consists of the Gender, Eye colour, Hair colour, and Subject height elements, the Properties block, the Expression block, and the Pose angle block. 內容:身分中繼資料區塊旨在描述影像中人物的屬性。身分中繼資料區塊包含性別、眼睛顏色、頭髮顏色和主體身高元素,以及屬性區塊、表情區塊和姿勢角度區塊。
If all elements of this element are absent, the element itself shall be absent, too. 如果此元素的所有子元素皆不存在,則此元素本身亦應不存在。
7.15 Gender 7.15 性別
Abstract values: The value of this element shall be one of the following: 抽象值:此元素的值應為下列其中之一:
unknown; 未知;
other; 其他;
male; 男;
female. 女。
Contents: The Gender element shall represent the gender of the subject. 內容:性別元素應代表受測者的性別。
7.16 Eye colour 7.16 眼睛顏色
Abstract values: The value of this element shall be one of the following: 抽象值:此元素的值應為下列其中之一:
unknown; 未知;
other; 其他;
black; 黑色;
blue; 藍色;
brown; 棕色;
grey; 灰色;
green; 綠色;
hazel; 淡褐色;
multi-coloured; 多色;
pink. 粉紅色。
Contents: The Eye colour element shall represent the colour of the irises of the eyes. If the eyes have different colours, then the colour of the right eye shall be encoded. 內容:眼睛顏色元素應表示虹膜的顏色。如果雙眼顏色不同,則應編碼右眼的顏色。
7.17 Hair colour 7.17 頭髮顏色
Abstract values: The value of this element shall be one of the following. 抽象值:此元素的值應為以下其中之一。
Contents: quad-quad\quad-\quad unknown; 內容: quad-quad\quad-\quad 未知;
other; 其他;
bald; 禿頭;
black; 黑髮;
blonde; 金髮;
brown; 棕髮;
grey; 灰色;
white; 白色;
red; 紅色;
known coloured (it is known that the hair colour has been changed from the natural one of that capture subject). 已知染色(已知髮色已從該擷取對象的自然髮色改變)。
Contents: The Hair colour element shall represent the colour of the hair of the subject. 內容:髮色元素應代表受測者的頭髮顏色。
7.18 Subject height 7.18 受測者身高
Abstract values: Integer. 抽象值:整數。
Contents: The Subject height element shall represent the height of the subject in millimetres. The minimum value for this element shall be 1 mm and the maximum value shall be 65535 mm . 內容:受測者身高元素應以公釐為單位表示受測者的身高。此元素的最小值應為 1 公釐,最大值應為 65535 公釐。
NOTE This value in most cases can only be used as a rough estimate of the subject height. Shoes, age, and even time of the day influence this measure. 註:此數值在大多數情況下只能作為受測者身高的粗略估計。鞋子、年齡,甚至一天中的時間都會影響此測量。
7.19 Properties block 7.19 屬性區塊
Abstract values: This element contains one or several of the following elements: 抽象值:此元素包含以下一個或多個元素:
glasses; 眼鏡;
moustache; 鬍子;
beard; 鬍鬚;
teeth visible; 牙齒可見;
pupil or iris not visible (e.g. either or both eyes closed or half closed); 瞳孔或虹膜不可見(例如,雙眼或單眼閉合或半閉合);
mouth open; 嘴巴張開;
left eye patch; 左眼眼罩;
right eye patch; 右眼眼罩;
dark glasses (medical); 深色眼鏡(醫療用);
biometric absent (conditions which could impact landmark detection); 生物特徵不存在(可能影響地標偵測的狀況);
head coverings present (e.g., hats, scarves, toupees). 頭部覆蓋物(例如:帽子、圍巾、假髮)是否存在。
Contents: The Properties block indicates which properties are present. There may be restrictions for different Face image kinds (see the profiles in Annex D). Each element may be true, false or absent. False elements do not need to be listed unless those elements are mandatory. 內容:屬性區塊表示存在哪些屬性。不同的臉部影像類型可能會有不同的限制(請參閱附錄 D 中的設定檔)。每個元素可以是真、假或不存在。除非是強制性元素,否則不需要列出為假的元素。
7.20 Expression block 7.20 表情區塊
Abstract values: This element contains one or several of the following items: 抽象值:此元素包含以下一項或多項:
neutral (non-smiling) with both eyes open and mouth closed; 面部表情中性(不微笑),雙眼睜開,嘴巴閉合;
smile; 微笑;
raised eyebrows; 眉毛上揚;
eyes looking away from the camera; 眼睛未注視鏡頭;
squinting; 瞇眼;
frowning. 皺眉。
Contents: The Expression block indicates which expressions are shown. Each element may be true, false or absent. False elements do not need to be listed unless those elements are mandatory. Neutral and smile shall not be both true for the same image. 內容:表情區塊表示顯示了哪些表情。每個元素可以是真、假或不存在。除非是強制性元素,否則不需要列出假元素。中性與微笑不能同時為真。
7.21 Pose angle block 7.21 姿勢角度區塊
Key 鍵
1 pitch § 1 俯仰 §
2 yaw (Y)(\mathrm{Y}) 2 偏擺 (Y)(\mathrm{Y}) 3roll(R)3 \operatorname{roll}(\mathrm{R})
The three elements together define the pose (Y, P, R). 這三個元素共同定義了姿態 (Y, P, R)。
Figure 3 - Definition of pose angles with respect to the frontal view of the subject 圖 3 - 受測者正面視圖的姿態角度定義
Abstract values: The Pose angle block contains Angle blocks for yaw, pitch, and roll. 抽象值:姿態角度區塊包含用於偏航、俯仰和滾轉的角度區塊。
Contents: The Pose angle block shall represent the estimated or measured pose of the subject in the image. 內容:姿態角度區塊應表示影像中主體的估計或測量姿態。
The angles encoded in this element are: 此元素中編碼的角度為:
Yaw angle block (Y): Rotation about the vertical (y) axis. The yaw angle Y is the rotation in degrees about the yy-axis (vertical axis) shown in Figure 3. Frontal poses have a yaw angle of 0^(@)0^{\circ}. Positive angles represent faces looking to their left (a counter-clockwise rotation around the yy-axis). 偏航角度區塊 (Y):繞垂直 (y) 軸的旋轉。偏航角度 Y 是繞圖 3 所示的 yy 軸(垂直軸)的度數旋轉。正面姿態的偏航角度為 0^(@)0^{\circ} 。正角度表示臉部朝向其左側(繞 yy 軸逆時針旋轉)。
Pitch angle block ( P ): Rotation about the horizontal side-to-side ( x ) axis. The pitch angle P is the rotation in degrees about the x -axis (horizontal axis) shown in Figure 3. Frontal poses have a pitch angle of 0^(@)0^{\circ}. Positive angles represent faces looking down (a counter-clockwise rotation around the x-axis). 俯仰角區塊 (P):繞水平左右 (x) 軸旋轉。俯仰角 P 是繞 x 軸(水平軸)的旋轉度數,如圖 3 所示。正面姿勢的俯仰角為 0^(@)0^{\circ} 。正角度表示臉部朝下(繞 x 軸逆時針旋轉)。
Roll angle block ®: Rotation about the horizontal back to front (z) axis. The roll angle R is the rotation in degrees about the z -axis (the horizontal axis from front to back) shown in Figure 3. Frontal poses have a roll angle of 0^(@)0^{\circ}. Positive angles represent faces tilted toward their right shoulder (counter-clockwise rotation around the z-axis). A roll angle of 0^(@)0^{\circ} denotes that the left and right eye centres have identical y coordinates. 滾轉角區塊 (R):繞水平前後 (z) 軸旋轉。滾轉角 R 是繞 z 軸(從前到後的水平軸)的旋轉度數,如圖 3 所示。正面姿勢的滾轉角為 0^(@)0^{\circ} 。正角度表示臉部向右肩傾斜(繞 z 軸逆時針旋轉)。滾轉角為 0^(@)0^{\circ} 表示左右眼中心具有相同的 y 座標。
The angles are defined relative to the frontal pose of the subject, which has angles (Y=P=R=0)(\mathrm{Y}=\mathrm{P}=\mathrm{R}=0) as shown in Figure 3. The frontal pose is defined by the Frankfurt Horizon as the xz plane and the vertical symmetry plane as the yz plane with the z axis oriented in the direction of the face sight. Examples are shown in Figure 4. 這些角度是相對於受測者的正面姿勢定義的,其角度為 (Y=P=R=0)(\mathrm{Y}=\mathrm{P}=\mathrm{R}=0) ,如圖 3 所示。正面姿勢由法蘭克福水平面定義為 xz 平面,垂直對稱平面定義為 yz 平面,z 軸指向臉部視線方向。範例如圖 4 所示。
As order of the successive rotation around the different axes does matter, the encoded rotation angle shall correspond to an order of execution starting from the frontal view. This order shall be given by roll (about the front axis), then pitch (about the horizontal axis) and finally yaw (about the vertical axis). The (first executed) roll transformation will therefore always be in the image xy plane. 由於繞不同軸的連續旋轉順序很重要,因此編碼的旋轉角度應對應於從正面視圖開始的執行順序。此順序應為滾轉(繞前軸),然後俯仰(繞水平軸),最後偏航(繞垂直軸)。因此,(首先執行的)滾轉變換將始終在影像 xy 平面中。
From the point of view of executing a transformation from the observed view to a frontal view, the transformation order will therefore be in the opposite order: Yaw, pitch, and then roll. The encoded angles are from the frontal view to the observed view. The conversion to integer is specified in 7.1. 從執行從觀察視圖到正面視圖的轉換角度來看,轉換順序將會是相反的:偏航、俯仰,然後是滾轉。編碼的角度是從正面視圖到觀察視圖。轉換為整數的規範在 7.1 中。
Figure 4 - Examples of pose angles in the form ( Y,P,R\mathrm{Y}, \mathrm{P}, \mathrm{R} ) 圖 4 - 姿態角度範例,形式為 ( Y,P,R\mathrm{Y}, \mathrm{P}, \mathrm{R} )
7.22 Angle data block 7.22 角度資料區塊
Abstract values: Angle value and Angle uncertainty. 抽象值:角度值和角度不確定性。
Contents: The Angle data block element contains an Angle value and its corresponding Angle uncertainty. 內容:角度資料區塊元素包含一個角度值及其對應的角度不確定性。
7.23 Angle value 7.23 角度值
Abstract values: Integer, the minimum value is -180 , the maximum value is 180 . 抽象值:整數,最小值為 -180,最大值為 180。
Contents: The Angle value is given by Tait-Bryan angles (in degrees). 內容:角度值由泰特-布萊恩角(以度為單位)給出。
7.24 Angle uncertainty 7.24 角度不確定性
Abstract values: The minimum value of an Angle uncertainty variable is 0 , the maximum value is 180 . 抽象值:角度不確定性變數的最小值為 0,最大值為 180。
Contents: The Angle uncertainty represents the expected degree of uncertainty of the associated pose angle. The more uncertain, the value of the uncertainty shall become larger. The Angle uncertainty allows storing an uncertainty or tolerance value for an angle. The true angle should be in a range of Angle value +-\pm Angle uncertainty. If the associated pose angle is absent, the Angle uncertainty for this angle shall be absent, too. 內容:角度不確定性表示相關聯的姿態角度的預期不確定程度。不確定性越大,不確定性的值應越大。角度不確定性允許儲存角度的不確定性或容許值。真實角度應在角度值 +-\pm 角度不確定性的範圍內。如果相關聯的姿態角度不存在,則該角度的角度不確定性也應不存在。
7.25 Landmark block 7.25 地標區塊
Abstract values: None. 抽象值:無。
Contents: The Landmark block specifies the type, code and position of landmarks in the face image. If the Landmark blocks element is present, it shall contain at least one Landmark block. A Landmark block consists of the Landmark kind element and the Landmark coordinates block. The structure of this element is shown in Figure 1. 目錄:地標區塊指定臉部影像中地標的類型、代碼和位置。如果地標區塊元素存在,則應至少包含一個地標區塊。地標區塊由地標種類元素和地標座標區塊組成。此元素的結構如圖 1 所示。
Landmarks can be specified as MPEG-4 feature points as given by ISO/IEC 14496-2:2004, Annex C or as anthropometric landmarks. The description of the anthropometric landmarks ^([2]){ }^{[2]} and their relation with the set of MPEG4 feature points is discussed in Table 2. 地標可指定為 ISO/IEC 14496-2:2004 附錄 C 中所定義的 MPEG-4 特徵點,或指定為人體測量地標。人體測量地標 ^([2]){ }^{[2]} 的描述及其與 MPEG4 特徵點集的關係將在表 2 中討論。
7.26 Landmark kind 7.26 地標種類
Abstract values: MPEG4 feature point or anthropometric landmark. 抽象值:MPEG4 特徵點或人體測量地標。
Contents: The Landmark kind shall either be MPEG4 feature point or anthropometric landmark. The Landmark code shall specify the landmark that is stored in the Landmark block element. The MPEG4 feature points are extended by eye and nostril landmarks 內容:地標類型應為 MPEG4 特徵點或人體測量學地標。地標代碼應指定儲存在地標區塊元素中的地標。MPEG4 特徵點透過眼睛和鼻孔地標進行擴展。
References to right and left shall be taken from the perspective of the subject contained within an image. References to right shall mean the right side of the body from the perspective of the subject. References to the left shall mean the left side of the body from the perspective of the subject. 左右的參考應從影像中主體的視角來看。右的參考應指從主體視角來看身體的右側。左的參考應指從主體視角來看身體的左側。
7.27 MPEG4 feature point 7.27 MPEG4 特徵點
Abstract values: See Figure 5 and Figure 6. 抽象值:請參閱圖 5 和圖 6。
Contents: Figure 5 denotes the landmark codes associated with feature points as given by ISO/ IEC 14496-2:2004, Annex C. Each landmark can be written in the form A.B using a major (A)(\mathrm{A}) and a minor (B)(\mathrm{B}) value. Eye and nostril landmarks are contained as an addition to the MPEG4 feature points. 內容:圖 5 所示為 ISO/IEC 14496-2:2004 附錄 C 中所定義的與特徵點相關的地標代碼。每個地標都可以用 A.B 的形式表示,其中 A 為主要值 (A)(\mathrm{A}) ,B 為次要值 (B)(\mathrm{B}) 。眼睛和鼻孔的地標是作為 MPEG4 特徵點的補充而包含的。
Key 圖例
feature points affected by face animation parameters (FAPs) as specified in ISO/IEC 14496-2 受 ISO/IEC 14496-2 中所定義的臉部動畫參數 (FAP) 影響的特徵點
other feature points 其他特徵點
Figure 5 - Feature points as specified in ISO/IEC 14496-2 圖 5 - ISO/IEC 14496-2 中所定義的特徵點
The eye centre landmarks 12.1 (left) and 12.2 (right) are defined to be the horizontal and vertical midpoints of the eye corners (3.7, 3.11) and (3.8, 3.12) respectively. The left nostril centre landmark 12.3 is defined to be the midpoint of the nose landmarks ( 9.1,9.159.1,9.15 ) in the horizontal direction and ( 9.3,9.159.3,9.15 ) in the vertical direction. Similarly, the right nostril centre landmark 12.4 is defined to be the midpoint of the nose landmarks ( 9.2,9.159.2,9.15 ) in the horizontal direction and ( 9.3,9.159.3,9.15 ) in the vertical direction. Both the eye centre and nostril centre landmarks are shown in Figure 6 and values given in Table 1. 眼睛中心地標 12.1(左)和 12.2(右)分別定義為眼角(3.7、3.11)和(3.8、3.12)的水平和垂直中點。左鼻孔中心地標 12.3 定義為鼻部地標在水平方向的( 9.1,9.159.1,9.15 )和垂直方向的( 9.3,9.159.3,9.15 )中點。同樣地,右鼻孔中心地標 12.4 定義為鼻部地標在水平方向的( 9.2,9.159.2,9.15 )和垂直方向的( 9.3,9.159.3,9.15 )中點。眼睛中心和鼻孔中心地標均顯示於圖 6,其數值列於表 1。
Key 圖例
feature points affected by FAPs 受 FAP 影響的特徵點
other feature points 其他特徵點
The landmarks 12.1, 12.2, 12.3, and 12.4 are defined to be the midpoints of MPEG features. 地標 12.1、12.2、12.3 和 12.4 定義為 MPEG 特徵的中點。
Figure 6 - Eye and nostril centre landmarks 圖 6 - 眼睛和鼻孔中心地標
Table 1 - Eye and nostril centre landmark codes 表 1 - 眼睛和鼻孔中心地標代碼
Centre landmark 中心地標
Midpoint of landmarks 地標中點
Landmark code 地標法典
Left eye 左眼
3.7,3.113.7,3.11
12.1
Right eye 右眼
3.8,3.123.8,3.12
12.2
Left nostril 左鼻孔
Horizontal 水準
Vertical 垂直
12.3
9.1,9.159.1,9.15
9.3,9.159.3,9.15
Horizontal 水準
Vertical 垂直
12.4
9.2,9.159.2,9.15
9.3,9.159.3,9.15
Centre landmark Midpoint of landmarks Landmark code
Left eye 3.7,3.11 12.1
Right eye 3.8,3.12 12.2
Left nostril Horizontal Vertical 12.3
9.1,9.15 9.3,9.15
Horizontal Vertical 12.4
9.2,9.15 9.3,9.15 | Centre landmark | Midpoint of landmarks | | Landmark code |
| :--- | :--- | :--- | :--- |
| Left eye | $3.7,3.11$ | 12.1 | |
| Right eye | $3.8,3.12$ | 12.2 | |
| Left nostril | Horizontal | Vertical | 12.3 |
| | $9.1,9.15$ | $9.3,9.15$ | |
| | Horizontal | Vertical | 12.4 |
| | $9.2,9.15$ | $9.3,9.15$ | |
ISO/IEC 39794-5:2019(E)
7.28 Anthropometric landmark 7.28 人體測量學地標
Abstract values: See Table 2. 抽象值:請參閱表 2。
Contents: Anthropometric landmarks denote feature points that are used in forensics and anthropology for recognition of individuals via two face images or image and skull over a long time. They also allow specification of points that are in use by criminal examiners and anthropologists ^([2]){ }^{[2]}. 內容:人體測量學地標表示特徵點,這些特徵點在法醫學和人類學中用於透過兩張臉部影像或影像與顱骨在長時間內識別個體。它們也允許指定犯罪檢查員和人類學家正在使用的點 ^([2]){ }^{[2]} 。
Figure 7 and Table 2 show the definition of the Anthropometric landmarks. The set of points represents the craniofacial landmarks of the head and face. The latter are used in forensics for “face to face” and “skull to face” identification. Some of these points have MPEG 4 counterparts, others not. 圖 7 和表 2 顯示了人體測量學地標的定義。這些點集代表頭部和臉部的顱顏地標。後者在法醫學中用於「面對面」和「顱骨對臉部」的識別。其中一些點有 MPEG 4 對應點,而另一些則沒有。
There are three different possibilities to encode an Anthropometric landmark: 有三種不同的方式可以編碼人體測量學地標:
Firstly, each Anthropometric landmark may be notated in the form A.B. A specifies the global landmark of the face to which this landmark belongs such as nose, mouth, etc. B specifies the particular point. In case a landmark has two symmetrical entities (left and right) the right entity always has a greater and even minor code value. Hence, all landmarks from the left part of the face have odd minor codes, and from the right part - even minor codes. 首先,每個人體測量學地標可以 A.B 的形式標記。A 指定該地標所屬的臉部整體地標,例如鼻子、嘴巴等。B 指定特定點。如果一個地標有兩個對稱實體(左和右),則右實體總是具有較大且偶數的次要代碼值。因此,臉部左側的所有地標都具有奇數次要代碼,而右側則具有偶數次要代碼。
Secondly, each Anthropometric landmark may be notated by its name. In case a landmark has two symmetrical entities (left and right) a “left” or “right” shall be added to the names in Table 2. 其次,每個人體測量學地標可以其名稱標記。如果一個地標有兩個對稱實體(左和右),則應在表 2 中的名稱後加上「左」或「右」。
Thirdly, each Anthropometric landmark may be notated by its point identifier. In case a landmark has two symmetrical entities (left and right) a “left” or “right” shall be added to the names in Table 2. 第三,每個人體測量學地標可以其點識別碼標記。如果一個地標有兩個對稱實體(左和右),則應在表 2 中的名稱後加上「左」或「右」。
Key 圖例
landmarks without MPEG4 counterpart 無 MPEG4 對應項之地標
landmarks with MPEG4 counterpart 有 MPEG4 對應項之地標
Figure 7 - Anthropometric landmarks 圖 7 - 人體測量學地標
Table 2 - Definitions of the anthropometric landmarks 表 2 - 人體測量學地標之定義
Anthropometric landmark point identifier 人體測量學地標點識別碼
Anthropometric landmark point name 人體測量學地標點名稱
MPEG4
Anthropometric landmark name 人體測量學地標名稱
How to point 如何標記
v
1.1
11.4
vertex 頂點
The highest point of head when the head is oriented in Frankfurt Horizon 當頭部朝向法蘭克福水平線時,頭部的最高點
g
1.2
glabella 眉心
The most prominent middle point between the eyebrows 眉毛之間最突出的中點
op 枕骨點
1.3
opisthocranion 後顱點
Situated in the occipital region of the head is most distant from the glabella 位於頭部枕骨區域,距離眉心最遠的點
eu 歐
1.5, 1.6
eurion 耳點
The most prominent lateral point on each side of the skull in the area of the parietal and temporal bones 位於顱骨兩側,頂骨和顳骨區域最突出的外側點
ft 額顳點
1.7, 1.8
frontotemporale 前顳點
The point on each side of the forehead, laterally from the elevation of the linea temporalis 額頭兩側,顳線隆起處外側的點
tr 髮際點
1.9
11.1
trichion 髮際點
The point on the hairline in the midline of the forehead 額頭中線髮際線上的點
zy 顴點
2.1, 2.2
zygion 顴點
The most lateral point of each of the zygomatic 顴骨兩側最外側的點
go 下顎角點
2.3,2.4
2.13, 2.14
gonion 下顎角點
The most lateral point on the mandibural angle close to the bony gonion 下顎角最外側點,靠近骨性下顎角
Anthropometric landmark point identifier Anthropometric landmark point name MPEG4 Anthropometric landmark name How to point
v 1.1 11.4 vertex The highest point of head when the head is oriented in Frankfurt Horizon
g 1.2 glabella The most prominent middle point between the eyebrows
op 1.3 opisthocranion Situated in the occipital region of the head is most distant from the glabella
eu 1.5, 1.6 eurion The most prominent lateral point on each side of the skull in the area of the parietal and temporal bones
ft 1.7, 1.8 frontotemporale The point on each side of the forehead, laterally from the elevation of the linea temporalis
tr 1.9 11.1 trichion The point on the hairline in the midline of the forehead
zy 2.1, 2.2 zygion The most lateral point of each of the zygomatic
go 2.3,2.4 2.13, 2.14 gonion The most lateral point on the mandibural angle close to the bony gonion| Anthropometric landmark point identifier | Anthropometric landmark point name | MPEG4 | Anthropometric landmark name | How to point |
| :--- | :--- | :--- | :--- | :--- |
| v | 1.1 | 11.4 | vertex | The highest point of head when the head is oriented in Frankfurt Horizon |
| g | 1.2 | | glabella | The most prominent middle point between the eyebrows |
| op | 1.3 | | opisthocranion | Situated in the occipital region of the head is most distant from the glabella |
| eu | 1.5, 1.6 | | eurion | The most prominent lateral point on each side of the skull in the area of the parietal and temporal bones |
| ft | 1.7, 1.8 | | frontotemporale | The point on each side of the forehead, laterally from the elevation of the linea temporalis |
| tr | 1.9 | 11.1 | trichion | The point on the hairline in the midline of the forehead |
| zy | 2.1, 2.2 | | zygion | The most lateral point of each of the zygomatic |
| go | 2.3,2.4 | 2.13, 2.14 | gonion | The most lateral point on the mandibural angle close to the bony gonion |
ISO/IEC 39794-5:2019(E)
Table 2 (continued) 表 2 (續)
Anthropometric landmark point identifier 人體測量學地標點識別碼
Anthropometric landmark point name 人體測量學地標點名稱
MPEG4
Anthropometric landmark name 人體測量學地標名稱
How to point 如何標記
sl
2.5
sublabiale 唇下點
Determines the lower border of the lower lip or the upper border of the chin 決定下唇的下緣或下巴的上緣
pg 頦點
2.6
2.10
pogonion 頦點
The most anterior midpoint of the chin, located on the skin surface in the front of the identical bony landmark of the mandible 下巴最前方的中點,位於下頷骨相同骨性地標前方的皮膚表面
gn
2.7
2.1
menton (or gnathion) 頦點(或 gnathion)
The lowest median landmark on the lower border of the mandible 下頷骨下緣最低的中線地標
cdl
2.9, 2.10
condylion laterale
The most lateral point on the surface of the condyle of the mandible 下顎骨 condyle 表面最外側點
en
3.1, 3.2
3.11, 3.8
endocanthion 眼內眥
The point at the inner commissure of the eye fissure 眼裂內合點
ex 例如
3.3, 3.4
3.7, 3.12
exocanthion (or ectocanthion) 外眼角點(或外眥點)
The point at the outer commissure of the eye fissure 眼裂外側聯合處的點
p
3.5, 3.6
3.5, 3.6
centre point of pupil 瞳孔中心點
Is determined when the head is in the rest position and the eye is looking straight forward 在頭部處於靜止位置且眼睛直視前方時確定
or 或
3.7, 3.8
3.9, 3.10
orbitale 眼眶下緣點
The lowest point on the lower margin of each orbit 每個眼眶下緣的最低點
ps 眼瞼下緣點
3.9, 3.10
3.1, 3.2
palpebrale superius 眼瞼上緣點
The highest point in the midportion of the free margin of each upper eyelid 每個上眼瞼自由緣中段的最高點
pi
3.11, 3.12
3.3, 3.4
palpebrale inferius 下瞼緣
The lowest point in the midportion of the free margin of each lower eyelid 每隻下眼瞼自由緣中段的最低點
os
4.1, 4.2
orbitale superius
The highest point on the lower border of the eyebrow 眉毛下緣的最高點
sci
4.3, 4.4
4.3, 4.4
superciliare 眉上點
The highest point on the upper border in the midportion of each eyebrow 每隻眉毛中段上緣的最高點
n
5.1
nasion 鼻根點
The point in the middle of both the nasal root and nasofrontal suture 鼻根和鼻額縫合線中間的點
se 鼻中隔點
5.2
sellion (or subnasion) 鼻下點(或鼻下緣點)
Is the deepest landmark located on the bottom of the nasofrontal angle 位於鼻額角底部最深處的地標
al 翼點
5.3, 5.4
9.1, 9.2
alare 鼻翼點
The most lateral point on each alar contour 每個鼻翼輪廓最外側的點
prn 鼻尖點
5.6
9.3
pronasale 鼻尖點
The most protruded point of the apex nasi 鼻尖最突出的點
sn 鼻下點
9.15
subnasale 鼻下點
The craniometric point at the base of the nasal (nose) spine 鼻棘底部的人體測量點
sbal
subalare
ac
9.1, 9.2
alar curvature (or alar crest) point 鼻翼曲率(或鼻翼嵴)點
The nasal alar crest 鼻翼脊
mf
9.6, 9.7
maxillofrontale 上頜額縫點
cph
8.9, 8.10
christa philtra landmark 人中脊點
The point on the crest of the philtrum, the vertical groove in the median portion of the upper lip, just above the vermillion border (sharp demarcation between the lip and the adjacent normal skin) 人中脊上的點,即上唇中線部位的垂直凹槽,位於唇紅緣(嘴唇與相鄰正常皮膚之間的明顯分界線)正上方
Anthropometric landmark point identifier Anthropometric landmark point name MPEG4 Anthropometric landmark name How to point
sl 2.5 sublabiale Determines the lower border of the lower lip or the upper border of the chin
pg 2.6 2.10 pogonion The most anterior midpoint of the chin, located on the skin surface in the front of the identical bony landmark of the mandible
gn 2.7 2.1 menton (or gnathion) The lowest median landmark on the lower border of the mandible
cdl 2.9, 2.10 condylion laterale The most lateral point on the surface of the condyle of the mandible
en 3.1, 3.2 3.11, 3.8 endocanthion The point at the inner commissure of the eye fissure
ex 3.3, 3.4 3.7, 3.12 exocanthion (or ectocanthion) The point at the outer commissure of the eye fissure
p 3.5, 3.6 3.5, 3.6 centre point of pupil Is determined when the head is in the rest position and the eye is looking straight forward
or 3.7, 3.8 3.9, 3.10 orbitale The lowest point on the lower margin of each orbit
ps 3.9, 3.10 3.1, 3.2 palpebrale superius The highest point in the midportion of the free margin of each upper eyelid
pi 3.11, 3.12 3.3, 3.4 palpebrale inferius The lowest point in the midportion of the free margin of each lower eyelid
os 4.1, 4.2 orbitale superius The highest point on the lower border of the eyebrow
sci 4.3, 4.4 4.3, 4.4 superciliare The highest point on the upper border in the midportion of each eyebrow
n 5.1 nasion The point in the middle of both the nasal root and nasofrontal suture
se 5.2 sellion (or subnasion) Is the deepest landmark located on the bottom of the nasofrontal angle
al 5.3, 5.4 9.1, 9.2 alare The most lateral point on each alar contour
prn 5.6 9.3 pronasale The most protruded point of the apex nasi
sn 9.15 subnasale The craniometric point at the base of the nasal (nose) spine
sbal subalare
ac 9.1, 9.2 alar curvature (or alar crest) point The nasal alar crest
mf 9.6, 9.7 maxillofrontale
cph 8.9, 8.10 christa philtra landmark The point on the crest of the philtrum, the vertical groove in the median portion of the upper lip, just above the vermillion border (sharp demarcation between the lip and the adjacent normal skin)| Anthropometric landmark point identifier | Anthropometric landmark point name | MPEG4 | Anthropometric landmark name | How to point |
| :--- | :--- | :--- | :--- | :--- |
| sl | 2.5 | | sublabiale | Determines the lower border of the lower lip or the upper border of the chin |
| pg | 2.6 | 2.10 | pogonion | The most anterior midpoint of the chin, located on the skin surface in the front of the identical bony landmark of the mandible |
| gn | 2.7 | 2.1 | menton (or gnathion) | The lowest median landmark on the lower border of the mandible |
| cdl | 2.9, 2.10 | | condylion laterale | The most lateral point on the surface of the condyle of the mandible |
| en | 3.1, 3.2 | 3.11, 3.8 | endocanthion | The point at the inner commissure of the eye fissure |
| ex | 3.3, 3.4 | 3.7, 3.12 | exocanthion (or ectocanthion) | The point at the outer commissure of the eye fissure |
| p | 3.5, 3.6 | 3.5, 3.6 | centre point of pupil | Is determined when the head is in the rest position and the eye is looking straight forward |
| or | 3.7, 3.8 | 3.9, 3.10 | orbitale | The lowest point on the lower margin of each orbit |
| ps | 3.9, 3.10 | 3.1, 3.2 | palpebrale superius | The highest point in the midportion of the free margin of each upper eyelid |
| pi | 3.11, 3.12 | 3.3, 3.4 | palpebrale inferius | The lowest point in the midportion of the free margin of each lower eyelid |
| os | 4.1, 4.2 | | orbitale superius | The highest point on the lower border of the eyebrow |
| sci | 4.3, 4.4 | 4.3, 4.4 | superciliare | The highest point on the upper border in the midportion of each eyebrow |
| n | 5.1 | | nasion | The point in the middle of both the nasal root and nasofrontal suture |
| se | 5.2 | | sellion (or subnasion) | Is the deepest landmark located on the bottom of the nasofrontal angle |
| al | 5.3, 5.4 | 9.1, 9.2 | alare | The most lateral point on each alar contour |
| prn | 5.6 | 9.3 | pronasale | The most protruded point of the apex nasi |
| sn | | 9.15 | subnasale | The craniometric point at the base of the nasal (nose) spine |
| sbal | | | subalare | |
| ac | | 9.1, 9.2 | alar curvature (or alar crest) point | The nasal alar crest |
| mf | | 9.6, 9.7 | maxillofrontale | |
| cph | | 8.9, 8.10 | christa philtra landmark | The point on the crest of the philtrum, the vertical groove in the median portion of the upper lip, just above the vermillion border (sharp demarcation between the lip and the adjacent normal skin) |
Table 2 (continued) 表 2(續)
Anthropometric landmark point identifier 人體測量學地標點識別碼
Anthropometric landmark point name 人體測量學地標點名稱
MPEG4
Anthropometric landmark name 人體測量學地標名稱
How to point 如何指向
ls
8.1
labiale (or labrale) superius 上唇點
The mid point of the vermillion border of the upper lip 上唇唇紅緣中點
li 下唇緣中點
8.2
labiale (or labrale) inferius 下唇唇紅緣中點
The mid point of the vermillion border of the lower lip 下唇唇紅緣中點
ch 唇角
8.3, 8.4
cheilion 唇角
The outer corner of the mouth where the outer edges of the upper and lower vermillions meet 口角,即上唇和下唇唇紅緣外側邊緣的交會處
sto
stomion
The median point of the oral slit when the lips are closed 當嘴唇閉合時,口腔裂縫的中點
sa
10.1, 10.2
superaurale 耳上點
The furthermost point of the ear lobe when measured from the sba landmark 從 sba 地標測量時,耳垂最遠的點
sba 耳下點
10.5, 10.6
subaurale 耳下點
The lowest point on the inferior (lower) border of the ear lobule when the subject is looking straight ahead 當受試者直視前方時,耳垂下緣(下方)的最低點
pra 耳前點
10.9, 10.10
preaurale 耳前點
The point between obs and abi opposite to pa 位於 obs 和 abi 之間,與 pa 相對的點
pa 耳後點
postaurale 耳後點
The most posterior point on the free margin of the ear 耳朵自由邊緣最靠後點
obs
10.3, 10.4
otobasion superius 耳上點
The highest point of attachment of the external ear to the head 外耳與頭部連接的最高點
obi 耳下點
obotasion inferius 耳下點
The lowest point of attachment of the external ear to the head 外耳與頭部連接的最低點
po 耳點
porion (soft) 耳點(軟)
The central point on the upper margin of the external auditory meatus (passage in the ear) 外耳道(耳朵中的通道)上緣的中心點
t
tragion
A cephalometric point in the notch just above the tragus (small tonguelike projection of the auricular cartilage) of the ear 耳屏(耳軟骨的小舌狀突起)正上方凹口處的頭顱測量點
Anthropometric landmark point identifier Anthropometric landmark point name MPEG4 Anthropometric landmark name How to point
ls 8.1 labiale (or labrale) superius The mid point of the vermillion border of the upper lip
li 8.2 labiale (or labrale) inferius The mid point of the vermillion border of the lower lip
ch 8.3, 8.4 cheilion The outer corner of the mouth where the outer edges of the upper and lower vermillions meet
sto stomion The median point of the oral slit when the lips are closed
sa 10.1, 10.2 superaurale The furthermost point of the ear lobe when measured from the sba landmark
sba 10.5, 10.6 subaurale The lowest point on the inferior (lower) border of the ear lobule when the subject is looking straight ahead
pra 10.9, 10.10 preaurale The point between obs and abi opposite to pa
pa postaurale The most posterior point on the free margin of the ear
obs 10.3, 10.4 otobasion superius The highest point of attachment of the external ear to the head
obi obotasion inferius The lowest point of attachment of the external ear to the head
po porion (soft) The central point on the upper margin of the external auditory meatus (passage in the ear)
t tragion A cephalometric point in the notch just above the tragus (small tonguelike projection of the auricular cartilage) of the ear| Anthropometric landmark point identifier | Anthropometric landmark point name | MPEG4 | Anthropometric landmark name | How to point |
| :--- | :--- | :--- | :--- | :--- |
| ls | | 8.1 | labiale (or labrale) superius | The mid point of the vermillion border of the upper lip |
| li | | 8.2 | labiale (or labrale) inferius | The mid point of the vermillion border of the lower lip |
| ch | | 8.3, 8.4 | cheilion | The outer corner of the mouth where the outer edges of the upper and lower vermillions meet |
| sto | | | stomion | The median point of the oral slit when the lips are closed |
| sa | | 10.1, 10.2 | superaurale | The furthermost point of the ear lobe when measured from the sba landmark |
| sba | | 10.5, 10.6 | subaurale | The lowest point on the inferior (lower) border of the ear lobule when the subject is looking straight ahead |
| pra | | 10.9, 10.10 | preaurale | The point between obs and abi opposite to pa |
| pa | | | postaurale | The most posterior point on the free margin of the ear |
| obs | | 10.3, 10.4 | otobasion superius | The highest point of attachment of the external ear to the head |
| obi | | | obotasion inferius | The lowest point of attachment of the external ear to the head |
| po | | | porion (soft) | The central point on the upper margin of the external auditory meatus (passage in the ear) |
| t | | | tragion | A cephalometric point in the notch just above the tragus (small tonguelike projection of the auricular cartilage) of the ear |
7.29 Landmark coordinates block 7.29 地標座標區塊
Abstract values: None. 抽象值:無。
Contents: The Landmark coordinates block shall contain the coordinates of the associated landmark in the 2D Cartesian coordinate system (in case of 2D image representation block existence), in a Coordinate texture image block, or in a 3D Cartesian coordinate system (in case of 3D image representation block existence). 內容:地標座標區塊應包含相關地標在 2D 笛卡爾座標系統(若存在 2D 影像表示區塊)、座標紋理影像區塊或 3D 笛卡爾座標系統(若存在 3D 影像表示區塊)中的座標。
In 2D Image representation blocks, the Z coordinate of the Cartesian coordinate system is not used. This element shall then contain the horizontal and vertical position of the associated landmark. They are measured in pixels with values from 0 to width- 1 and from 0 to height-1. The Coordinate texture image block consists of the two integer values uInPixel and vInPixel. In 3D Shape representation blocks, the X, Y, and Z coordinates are mandatory and defined in the 3D Cartesian coordinate system. The X,YX, Y, and Z coordinates are non-negative integers. The landmarks are converted to metric Cartesian coordinates using the Cartesian scales and offsets block. The error of the Z coordinate of an anthropometric landmark location should be no greater than 3 mm . The point shall withstand from the nearest point on the surface no further than 3 mm . 在 2D 影像表示區塊中,不使用笛卡爾座標系統的 Z 座標。此元素應包含相關地標的水平和垂直位置。它們以像素為單位測量,數值範圍為 0 到寬度-1 以及 0 到高度-1。座標紋理影像區塊由兩個整數值 uInPixel 和 vInPixel 組成。在 3D 形狀表示區塊中,X、Y 和 Z 座標是強制性的,並在 3D 笛卡爾座標系統中定義。 X,YX, Y 和 Z 座標是非負整數。地標使用笛卡爾比例和偏移區塊轉換為公制笛卡爾座標。人體測量地標位置的 Z 座標誤差應不超過 3 毫米。該點與表面最近點的距離應不超過 3 毫米。
7.30 Image representation block 7.30 影像表示區塊
Abstract values: Either 2D image representation block, or 3D shape representation block. 抽象值:2D 影像表示區塊,或 3D 形狀表示區塊。
Contents: The Image representation block contains the image data and metadata. It is either a 2D image representation block or a 3D shape representation block. 內容:影像表示區塊包含影像資料和中繼資料。其為 2D 影像表示區塊或 3D 形狀表示區塊。
7.31 2D image representation block 7.31 二維影像表示區塊
Abstract values: None. 抽象值:無。
Contents: The 2D image representation block contains the 2D representation data, the 2D image information block, and the 2D capture device block. 內容:2D 影像表示區塊包含 2D 表示資料、2D 影像資訊區塊和 2D 擷取裝置區塊。
7.32 2D representation data 7.32 二維表示資料
Abstract values: Octet string. 抽象值:八位元字串。
Contents: The 2D representation data element shall contain the encoded image data in accordance with the value of the Image data format element. 內容:2D 表示資料元素應包含依據影像資料格式元素值編碼的影像資料。
7.33 2D capture device block 7.33 二維擷取裝置區塊
Abstract values: None. 抽象值:無。
Contents: The 2D capture device block consists of the 2D capture device spectral block and the 2D capture device technology identifier. 內容:2D 擷取裝置區塊由 2D 擷取裝置光譜區塊和 2D 擷取裝置技術識別碼組成。
Abstract values: The possible values are: 抽象值:可能的值為:
near infrared; 近紅外線;
thermal; 熱感;
white light. 白光。
Contents: Many different types of capture devices work in the near infrared, thermal, or white light spectral range. The 2D capture device spectral block indicates whether the capture device technology uses one or more of these spectral ranges. 內容:許多不同類型的擷取裝置在近紅外線、熱或白光光譜範圍內運作。2D 擷取裝置光譜區塊指示擷取裝置技術是否使用這些光譜範圍中的一個或多個。
Abstract values: The possible values are: 抽象值:可能的值為:
unknown; 未知;
static photograph from an unknown source; 來自未知來源的靜態照片;
static photograph from a digital still-image camera; 來自數位相機的靜態照片;
static photograph from a scanner; 來自掃描器的靜態照片;
video frame(s) from an unknown source; 來自未知來源的視訊影格;
video frame(s) from an analogue video camera; 來自類比攝影機的視訊影格;
video frame(s) from a digital video camera. 來自數位攝影機的視訊影格。
Contents: The 2D capture device technology identifier shall indicate the device technology used to acquire the captured biometric sample. 內容:2D 擷取裝置技術識別碼應指出用於擷取生物識別樣本的裝置技術。
7.36 2D image information block 7.36 2D 影像資訊區塊
Abstract values: None. 抽象值:無。
Contents: The 2D image information block element is intended to describe digital properties of the 2D representation data. 目錄:2D 影像資訊區塊元素旨在描述 2D 表示資料的數位屬性。
The 2D image information block consists of the Image data format, the 2D face image kind, the Post-acquisition processing block, the Lossy transformation attempts element, the Camera to subject distance, the Sensor diagonal, the Lens focal length, the Image size block, the Image face measurements block, the Image colour space element, and the Reference colour mapping block. The structure of this element is shown in Figure 1. 2D 影像資訊區塊包含影像資料格式、2D 臉部影像類型、後處理區塊、有損轉換嘗試元素、相機到主體的距離、感測器對角線、鏡頭焦距、影像尺寸區塊、影像臉部測量區塊、影像色彩空間元素以及參考色彩映射區塊。此元素的結構如圖 1 所示。
7.37 2D face image kind 7.37 2D 臉部影像類型
Abstract values: See Table 3 for a list of allowed 2D face image kinds and their normative requirements. Other application specific image types may be added in the future. 抽象值:允許的 2D 臉部影像類型及其規範要求請參閱表 3。未來可能會新增其他應用程式特定的影像類型。
Contents: The 2D face image kind element shall represent the type of the face image stored in the 2D representation data. There are several types according to the chosen application specific profile (see Annex D), additional profiles may be included in future versions of this document. 內容:2D 臉部影像類型元素應表示儲存在 2D 表示資料中的臉部影像類型。根據所選的應用程式特定個人資料(請參閱附錄 D),有數種類型,未來版本的文件中可能會包含額外的個人資料。
Table 3-2D face image kinds 表 3—2D 臉部影像種類
Value 數值
Definition and normative requirements 定義與規範要求
MRTD
Annex D.1 附錄 D.1
General purpose 一般用途
Annex D.2 附錄 D.2
Value Definition and normative requirements
MRTD Annex D.1
General purpose Annex D.2| Value | Definition and normative requirements |
| :--- | :--- |
| MRTD | Annex D.1 |
| General purpose | Annex D.2 |
7.38 Post acquisition processing block 7.38 後處理區塊
Abstract values: The values of this block shall be one or more of the following: 抽象值:此區塊的值應為以下其中一個或多個:
rotated (in- plane); 旋轉(平面內);
cropped; 裁切;
down-sampled; 降取樣;
white balance adjusted; 白平衡已調整;
multiply compressed; 多重壓縮;
interpolated; 內插;
contrast stretched; 對比度拉伸;
pose corrected; 姿勢校正;
multi view image; 多視角影像;
age progressed; 年齡增長;
super-resolution processed; 超解析度處理;
normalised. 正規化。
There may be restrictions on the allowed values by the choice of the 2D face image kind. 允許值可能會因選擇的 2D 臉部影像類型而受到限制。
Contents: This element contains notifications on potential post acquisition processing steps. 內容:此元素包含有關潛在後續處理步驟的通知。
While the alteration of face image data is discouraged, there are cases when no alternative may exist: 雖然不鼓勵更改臉部影像資料,但在某些情況下可能沒有其他替代方案:
Legacy database of 3//43 / 4 frontal face images which shall be rotated to full frontal prior to biometric comparison. 包含 3//43 / 4 張正面臉部影像的舊有資料庫,在進行生物識別比對前應將其旋轉為完全正面。
Froma frontal image artificial non-frontal face images are automatically generated at predetermined non-frontal poses (multi-view images) using an implicit head model or similar. These images can be beneficial during the comparison process or a manual review process as they show a more similar pose than the original frontal image. 從正面影像中,使用隱式頭部模型或類似方法,在預定的非正面姿勢(多視角影像)下自動生成人工非正面臉部影像。這些影像在比對過程或人工審查過程中可能會有益處,因為它們顯示的姿勢比原始正面影像更相似。
A single image is to be age progressed and used for verification of a passport holder. 單一影像將進行年齡增長處理,並用於護照持有人的驗證。
A short video stream is super-resolved to a single face image for comparison against a watch list. 短視訊串流經超解析度處理為單一臉部影像,以與觀察清單進行比對。
The Post acquisition processing block allows the specification of the kind of post processing that has been applied to the original captured image. 後續擷取處理區塊允許指定已應用於原始擷取影像的後續處理類型。
On the one hand a captured image might need some post-processing so that the resulting representation conforms to the requirements of this document. On the other hand, these processing steps should be minimal and not distort the characteristics of the original image. 一方面,擷取到的影像可能需要一些後續處理,以使最終呈現符合本文件的要求。另一方面,這些處理步驟應盡量減少,且不應扭曲原始影像的特性。
7.39 Lossy transformation attempts 7.39 有損轉換嘗試
Abstract values: Unknown, 0,1 , more than 1 . 抽象值:未知、0、1、大於 1。
Contents: This element counts the number of previous lossy transformation steps. 內容:此元素計算先前有損轉換步驟的數量。
7.40 Image data format 7.40 影像資料格式
Abstract values: The values shall be specified according to Table 4. 抽象值:這些值應根據表 4 指定。
Contents: The Image data format denotes the encoding type of the 2D representation data and of the 3D texture map. 內容:影像資料格式表示 2D 表示資料和 3D 紋理貼圖的編碼類型。
For lossless compression PNG or JPEG2000 lossless shall be used. For lossless representation of images using more than 8 bits per channel PNG or JPEG2000 lossless shall be used. For lossy representation of images using more than eight bit per channel JPEG2000 shall be used. For an encoding in Netpbm portable binary the image formats P5 (grey, PGM) and P6 (colour, PPM) shall be used. 對於無損壓縮,應使用 PNG 或 JPEG2000 無損格式。對於每個通道超過 8 位元的影像無損表示,應使用 PNG 或 JPEG2000 無損格式。對於每個通道超過 8 位元的影像有損表示,應使用 JPEG2000 格式。對於 Netpbm 可攜式二進位編碼,應使用影像格式 P5(灰階,PGM)和 P6(彩色,PPM)。
Table 4 - Image data format codes 表 4 - 影像資料格式代碼
Value 數值
Specified in 指定於
unknown 未知
other 其他
jpeg
ITU-T Rec. T. 81 | ISO/IEC 10918-1 and Reference [3] ITU-T 建議書 T.81 | ISO/IEC 10918-1 和參考文獻 [3]
Value Specified in
unknown
other
jpeg ITU-T Rec. T. 81 | ISO/IEC 10918-1 and Reference [3]
jpeg2000 lossy jpeg2000 lossless ISO/IEC 15444-1
png ISO/IEC 15948
pgm Reference [33]
ppm Reference [34]| Value | Specified in |
| :--- | :--- |
| unknown | |
| other | |
| jpeg | ITU-T Rec. T. 81 \| ISO/IEC 10918-1 and Reference [3] |
| jpeg2000 lossy jpeg2000 lossless | ISO/IEC 15444-1 |
| png | ISO/IEC 15948 |
| pgm | Reference [33] |
| ppm | Reference [34] |
If the Image data format value is unknown or other or a later-version extension code, then the Image size block (with width and height) shall be included. 如果影像資料格式值為未知、其他或較新版本的擴充功能代碼,則應包含影像尺寸區塊(含寬度和高度)。
In the event that a greyscale face image is encoded in the Netpbm portable greyscale binary image format (PGM), the format definition is as follows: 如果灰階人臉影像以 Netpbm 可攜式灰階二進位影像格式 (PGM) 編碼,則其格式定義如下:
a “magic number” = “P5” for identifying the file type followed by: 一個「魔術數字」=「P5」,用於識別檔案類型,後接:
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
a width, formatted as ASCII characters in decimal; 寬度,以十進位 ASCII 字元格式化;
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
a height, formatted as ASCII characters in decimal; 高度,以十進位 ASCII 字元格式化;
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
the maximum grey value (Maxval), formatted as ASCII characters in decimal the value shall be smaller than 256, and larger than zero; 最大灰度值 (Maxval),以十進位 ASCII 字元格式表示,其值應小於 256 且大於零;
a single Whitespace character (usually a newline); 一個單一的空白字元(通常是換行符號);
a raster of Height rows, in order from top to bottom. Each row consists of Width grey values, in order from left to right. Each grey value is a number from 0 through Maxval, with 0 being black and Maxval being white. Each grey value is represented in pure binary by either 1 or 2 bytes. If the Maxval is less than 256, it is 1 byte. Otherwise, it is 2 bytes. The most significant byte is first. 一個由高 (Height) 列組成的點陣圖,順序從上到下。每一列包含寬 (Width) 個灰度值,順序從左到右。每個灰度值是一個從 0 到 Maxval 的數字,其中 0 為黑色,Maxval 為白色。每個灰度值以純二進位表示,佔用 1 或 2 個位元組。如果 Maxval 小於 256,則佔用 1 個位元組。否則,佔用 2 個位元組。最高有效位元組在前。
A PGM encoded greyscale face image shall be encoded in a P5 format. PGM 編碼的灰階人臉影像應以 P5 格式編碼。
In the event that a colour face image is encoded in the Netpbm portable colour binary image format (PPM), the format definition is as follows: 如果彩色人臉影像以 Netpbm 可攜式彩色二進位影像格式 (PPM) 編碼,其格式定義如下:
a “magic number” = “P6” for identifying the file type followed by: 一個「魔術數字」=「P6」,用於識別檔案類型,後接:
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
a width, formatted as ASCII characters in decimal; 一個寬度,以十進位 ASCII 字元格式表示;
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
a height, formatted as ASCII characters in decimal; 高度,以十進位 ASCII 字元格式表示;
any Whitespace (blanks, TABs, CRs, LFs); 任何空白字元(空格、TAB、CR、LF);
the maximum channel value (Maxval), formatted as ASCII characters in decimal - the value shall be smaller than 256, and larger than zero; 最大通道值(Maxval),以十進位 ASCII 字元格式表示——該值應小於 256 且大於零;
a single Whitespace character (usually a newline); 一個單一的空白字元(通常是換行符號);
a raster of Height rows, in order from top to bottom. Each row consists of Width pixel values, in order from left to right. Each pixel value is represented by 1 number for red, 1 number for green and 1 number for blue, each from 0 through Maxval; thus each pixel value is represented in pure binary by 3 bytes. 一個由高(Height)行組成的點陣圖,順序從上到下。每行由寬(Width)個像素值組成,順序從左到右。每個像素值由 1 個紅色數字、1 個綠色數字和 1 個藍色數字表示,每個數字的範圍從 0 到 Maxval;因此,每個像素值以純二進位表示為 3 個位元組。
A PPM encoded colour face image shall be encoded in a P6 format. PPM 編碼的彩色人臉影像應以 P6 格式編碼。
7.41 Camera to subject distance 7.41 相機到拍攝對象距離
Abstract values: Integer. 抽象值:整數。
Contents: The Camera to subject distance (CSD) element contains the camera to subject distance of the photographical setup used for capturing the photo in millimetres. The maximum CSD to be encoded is 50000 mm . All larger distances shall by encoded using that maximum value. 內容:相機到拍攝對象距離 (CSD) 元素包含用於拍攝照片的攝影裝置中相機到拍攝對象的距離,單位為毫米。可編碼的最大 CSD 為 50000 毫米。所有更大的距離應使用該最大值進行編碼。
7.42 Sensor diagonal 7.42 感測器對角線
Abstract values: Integer. 抽象值:整數。
Contents: The Sensor diagonal element contains the diagonal length of the camera sensor used for capturing the photo in millimetres. The maximum Sensor diagonal to be encoded is 2000 mm . All larger distances shall by encoded using that maximum value. If a zoom lens is used, this data element shall encode the actual focal length used to capture the image. 內容:感測器對角線元素包含用於拍攝照片的相機感測器的對角線長度,單位為毫米。可編碼的最大感測器對角線為 2000 毫米。所有更大的距離應使用該最大值進行編碼。如果使用變焦鏡頭,此資料元素應編碼用於拍攝影像的實際焦距。
Figure 8 illustrates the relative sizes of some commonly available image sensors. Table 5 provides the approximate widths, heights, areas, diagonals, and crop factors for these sensors. The dimensions in Table 5 are approximates and serve as examples. 圖 8 說明了一些常用影像感測器的相對尺寸。表 5 提供了這些感測器的約略寬度、高度、面積、對角線和裁切係數。表 5 中的尺寸為約略值,僅供參考。
Figure 8 - Typical sensors and their relation in size to the traditional full frame 圖 8 – 典型感測器及其與傳統全片幅的尺寸關係
It might be noted that, by gathering more light, a larger image sensor will provide typically lower image noise, and a fixed focal length lens will generally provide a higher image quality than a zoom lens of the same focal length. Moreover, by using a fixed focal length lens, the problem of inadvertent change to the zoom ratio (i.e., the field of view) can be avoided. 值得注意的是,較大的影像感測器透過收集更多光線,通常能提供較低的影像雜訊,而固定焦距鏡頭通常能提供比相同焦距的變焦鏡頭更高的影像品質。此外,使用固定焦距鏡頭可以避免無意中改變變焦比(即視場)的問題。
Abstract values: Integer. 抽象值:整數。
Contents: The Lens focal length element contains the focal length of the camera lens used for capturing the photo in millimetres. The maximum Lens focal length to be encoded is 2000 mm . All larger distances shall by encoded using that maximum value. 內容:鏡頭焦距元素包含用於拍攝照片的相機鏡頭焦距(以毫米為單位)。可編碼的最大鏡頭焦距為 2000 毫米。所有更大的距離應使用該最大值進行編碼。
7.44 Image size block 7.44 影像尺寸區塊
Abstract values: None. 抽象值:無。
Contents: The Image size block consists of the Width and the Height element. 內容:影像尺寸區塊由寬度和高度元素組成。
7.45 Width 7.45 寬度
Abstract values: Integer. 抽象值:整數。
Contents: The Width element shall specify the number of pixels of the 2D representation data in the horizontal direction. 內容:寬度元素應指定二維表示資料在水平方向的像素數量。
7.46 Height 7.46 高度
Abstract values: Integer. 抽象值:整數。
Contents: The Height element shall specify the number of pixels of the 2D representation data in the vertical direction. 內容:高度元素應指定二維表示資料在垂直方向的像素數量。
7.47 Image face measurements block 7.47 影像人臉測量區塊
Abstract values: None. 抽象值:無。
Contents: For specific application domains different minimal spatial sampling rates of the interchange data may be required. For example, using higher spatial sampling rate images allow for specific human as well as machine inspection methods that depend on the analysis of very small details. 內容:對於特定的應用領域,可能需要不同的交換資料最小空間取樣率。例如,使用較高空間取樣率的影像,可支援依賴於分析極微小細節的特定人工及機器檢測方法。
The Image face measurements block consists of four elements. If the number of pixels across the width of the head shall be encoded the Image head width may be used. If the number of pixels across the length of the head shall be encoded the Image head length may be used. If the inter-eye distance shall be encoded the Image inter-eye distance data element may be used. If the eye-to-mouth distance shall be encoded the Image eye-to-mouth distance data element may be used. If necessary, all four elements may be used. 影像臉部測量區塊包含四個元素。若要編碼頭部寬度的像素數,可使用「影像頭部寬度」。若要編碼頭部長度的像素數,可使用「影像頭部長度」。若要編碼眼距,可使用「影像眼距」資料元素。若要編碼眼到嘴距離,可使用「影像眼到嘴距離」資料元素。如有必要,這四個元素皆可使用。
7.48 Image head width 7.48 影像頭部寬度
Abstract values: Integer. 抽象值:整數。
Contents: The Image head width element provides information on the number of pixels in the image across the width of the head. The head width ( W ) is defined in Figure 9. 內容:影像頭部寬度元素提供頭部寬度方向的像素數量資訊。頭部寬度(W)定義於圖 9。
a) Abstract geometric characteristics a) 抽象幾何特性
b) Sample b) 樣本
Key 圖例
A image width 影像寬度 A
B image height 影像高度 B
W head width 頭部寬度 W
L head length 頭部長度 L
V vertical centre line V 垂直中心線
H horizontal centre line H 水平中心線
M face centre M 臉部中心
Figure 9 - Abstract geometric characteristics of a portrait applied to a sample 圖 9 - 應用於樣本的人像抽象幾何特性
NOTE The typical inter-eye distance is approximately half of the head width. 註:典型的雙眼間距約為頭部寬度的一半。
7.49 Image inter-eye distance 7.49 影像雙眼間距
Abstract values: Integer. 抽象值:整數。
Contents: The Image inter-eye distance element provides information on the number of pixels in the image between the eye centres (feature points 12.1 and 12.2). For an explanation of the inter-eye distance see Figure 10. The value of this element shall be the number of pixels between the eye centres. 內容:影像雙眼間距元素提供影像中雙眼中心(特徵點 12.1 和 12.2)之間像素數量的資訊。關於雙眼間距的說明請參閱圖 10。此元素的值應為雙眼中心之間的像素數量。
NOTE Be aware that the eye centre is not necessarily the centre of the pupil. 註:請注意,眼球中心不一定是瞳孔中心。
NOTE A typical real IED (distance measured at the face) is between 60 mm and 65 mm . 註:典型的實際 IED(臉部測量距離)介於 60 毫米至 65 毫米之間。
7.50 Image eye-to-mouth distance 7.50 影像眼口距離
Abstract values: Integer. 抽象值:整數。
Contents: The Image eye-to-mouth distance element provides information on the number of pixels in the image between the mouth and the eyes. The value of this element shall be the number of pixels between the midpoint of the line connecting the eye centres (feature points 12.1 and 12.2) and the mouth (feature point 2.3). 內容:影像眼到嘴距離元素提供影像中嘴巴和眼睛之間像素數量的資訊。此元素的值應為連接眼睛中心(特徵點 12.1 和 12.2)的線段中點與嘴巴(特徵點 2.3)之間的像素數量。
7.51 Image head length 7.51 影像頭部長度
Abstract values: Integer. 抽象值:整數。
Contents: The Image head length element provides information on the number of pixels in the image from the chin to crown, or length, of the head. The head length ( L ) is defined in Figure 9. The value of this element shall be the number of pixels across the length of the head. 內容:影像頭部長度元素提供影像中從下巴到頭頂的像素數量,即頭部的長度。頭部長度(L)定義於圖 9。此元素的值應為頭部長度方向的像素數量。
7.52 Image colour space 7.52 影像色彩空間
Abstract values: The value of this element shall be one of the following: 抽象值:此元素的值應為下列其中之一:
unknown; 未知;
other; 其他;
24 bit RGB; 24 位元 RGB;
48 bit RGB; 48 位元 RGB;
YUV422; YUV422;
8 bit greyscale; 8 位元灰階;
16 bit greyscale. 16 位元灰階。
Contents: The Image colour space element indicates the colour space used in the encoded 2D or 3D image information block. RGB encoding is recommended. The ICC profile should be embedded inside the Texture map data (if applicable), as JPEG and PNG formats allow ICC profile encoding. 內容:影像色彩空間元素表示編碼的 2D 或 3D 影像資訊區塊中使用的色彩空間。建議使用 RGB 編碼。ICC 設定檔應嵌入在紋理貼圖資料中(如果適用),因為 JPEG 和 PNG 格式允許 ICC 設定檔編碼。
7.53 Reference colour mapping block 7.53 參考色彩映射區塊
Abstract values: None. 抽象值:無。
Contents: Mapping of reference colours like in IEC 61966-8. This data element contains the name of the applied Reference colour schema, like IEC 61966-8, and a list of Reference colour definition and value blocks. 內容:參考色彩的映射,如 IEC 61966-8。此資料元素包含所應用參考色彩方案的名稱,如 IEC 61966-8,以及參考色彩定義和數值區塊的列表。
7.54 Reference colour schema 7.54 參考色彩架構
Abstract values: Octet string. 抽象值:八位元字串。
Contents: This data element contains the name of the applied Reference colour schema, like IEC 61966-8. 內容:此資料元素包含所應用參考色彩方案的名稱,如 IEC 61966-8。
7.55 Reference colour definition and value block 7.55 參考色彩定義和數值區塊
Abstract values: Two octet strings. 抽象值:兩個八位元字串。
Contents: These data elements contain pairs of elements consisting of a Reference colour definition like “J 14” in the IEC case, and the respective Reference colour value in the given face portrait. 內容:這些資料元素包含成對的元素,由參考色彩定義(例如 IEC 案例中的「J 14」)以及給定人臉肖像中的相應參考色彩值組成。
7.56 3D shape representation block 7.56 3D 形狀表示區塊
Abstract values: None. 抽象值:無。
Contents: The 3D shape representation block contains the 3D representation data, the 3D image information block, and the 3D capture device block. The structure of the 3D shape representation block is shown in Figure 1. 內容:3D 形狀表示區塊包含 3D 表示資料、3D 影像資訊區塊和 3D 擷取裝置區塊。3D 形狀表示區塊的結構如圖 1 所示。
7.57 3D representation data 7.57 3D 表示資料
Abstract values: Octet string. 抽象值:八位元字串。
Contents: The 3D representation data element shall contain the image data in a vertex representation. The 3D representation kind (vertex) shall be specified in the 3D representation kind element. 內容:3D 表示資料元素應包含頂點表示中的影像資料。3D 表示類型(頂點)應在 3D 表示類型元素中指定。
7.58 3D capture device block 7.58 3D 擷取裝置區塊
Abstract values: None. 抽象值:無。
Contents: In analogy to the 2D capture device block in the 2D image representation block, where the source of the 2D data can be coded, the 3D capture device block should be used to indicate the device that was used to acquire the 3D data. 內容:類似於 2D 影像表示區塊中的 2D 擷取裝置區塊,其中 2D 資料的來源可以編碼,3D 擷取裝置區塊應用於指示用於獲取 3D 資料的裝置。
The 3D capture device block consists of the 3D modus element and the 3D capture device technology identifier 3D element. 三維擷取裝置區塊由三維模式元素和三維擷取裝置技術識別碼三維元素組成。
If all elements of the 3D capture device block are absent the 3D capture device block element shall be absent. 如果 3D 擷取裝置區塊的所有元素都不存在,則 3D 擷取裝置區塊元素應不存在。
7.59 3D modus 7.59 3D 模式
Abstract values: The value of this element shall be one of the following: 抽象值:此元素的值應為以下其中之一:
unknown; 未知;
active; 啟用;
passive. 被動。
Contents: This element describes the manner in which the 3D image is acquired. 內容:此元素描述了擷取 3D 影像的方式。
7.60 3D capture device technology identifier 7.60 3D 擷取裝置技術識別碼
Abstract values: The value of this element shall be one of the following: 摘要值:此元素的數值應為下列其中之一:
unknown; 未知;
stereoscopic scanner; 立體掃描器;
moving (monochromatic) laser line; 移動式(單色)雷射線;
structured light; 結構光;
colour coded light; 色彩編碼光;
ToF (time of flight); 飛時測距(ToF);
shape from shading. 基於陰影的形狀重建。
Contents: This element contains information on the technology used in the capture device used. 內容:此元素包含所用擷取裝置中使用的技術資訊。
NOTE Some of the listed 3D capture device technology identifier abstract values are incompatible with a 3D modus value of passive. 註記:部分列出的 3D 擷取裝置技術識別碼抽象值與被動的 3D 模式值不相容。
7.61 3D image information block 7.61 3D 影像資訊區塊
Abstract values: None. 抽象值:無。
Contents: The 3D image information block consists of the the 3D representation kind block, the 3D coordinate system, the 3D Cartesian scales and offsets block, the Image colour space (see 7.52), the 3D face image kind, the Image size block (see 7.44), the 3D physical face measurements block, the Post acquisition processing block (see 7.38), and the 3D texture map block. The structure of this element is shown in Figure 1. 內容:3D 影像資訊區塊由 3D 表示種類區塊、3D 座標系統、3D 笛卡爾比例和偏移區塊、影像色彩空間(參見 7.52)、3D 臉部影像種類、影像尺寸區塊(參見 7.44)、3D 實體臉部測量區塊、擷取後處理區塊(參見 7.38)和 3D 紋理貼圖區塊組成。此元素的結構如圖 1 所示。
7.62 3D representation kind block 7.62 3D 表示種類區塊
Abstract values: Vertex. 抽象值:頂點。
Contents: The 3D representation kind block shall contain the name of the encoding schema used for the 3D representation data, which is 3D vertex block for this version of this document. 內容:3D 表示種類區塊應包含用於 3D 表示資料的編碼架構名稱,對於本文件的此版本而言,該名稱為 3D 頂點區塊。
3D vertex block codes 3D points based on a non-regular sampling interval, typically resulting in a sparse coding. Due to variable sampling of the vertex points the vertex representation on the one hand can result in very compact representations or in a very exact representation when using many vertices. 3D 頂點區塊根據非規則取樣間隔對 3D 點進行編碼,通常會產生稀疏編碼。由於頂點點的變數取樣,頂點表示一方面可以產生非常緊湊的表示,或者在使用許多頂點時產生非常精確的表示。
7.63 3D vertex block 7.63 3D 頂點區塊
Abstract values: None. 抽象值:無。
Contents: The 3D vertex block consists of at one or more 3D vertex information blocks, and one or more 3D vertex triangle data blocks. 內容:3D 頂點區塊由一個或多個 3D 頂點資訊區塊,以及一個或多個 3D 頂點三角形資料區塊組成。
The Coordinate system type for vertex data shall be Cartesian. All Cartesian coordinates shall be non-negative integer. After application of Cartesian scales and offsets, the Cartesian coordinates become metric Cartesian coordinates which can be negative and positive and decimal. 頂點資料的座標系統類型應為笛卡爾座標系。所有笛卡爾座標應為非負整數。在應用笛卡爾比例和偏移後,笛卡爾座標會變成公制笛卡爾座標,其可以是負數、正數和小數。
The origin of the metric Cartesian coordinates is defined. For example, this origin is linked to landmarks like the middle of the 2 eyes for the 3D textured image application profile, or like to the top of the nose. 公制笛卡爾座標的原點已定義。例如,此原點與地標相關聯,如用於 3D 紋理影像應用程式個人資料的兩眼之間,或與鼻尖相關聯。
The scale is defined to be in conformity with the 3D textured image resolution block. 比例的定義應符合 3D 紋理影像解析度區塊。
7.64 3D vertex information block 7.64 3D 頂點資訊區塊
Abstract values: None. 抽象值:無。
Contents: The 3D vertex information block consists of the 3D vertex coordinates block, the 3D vertex identifier, the 3D vertex normals block, the 3D vertex textures block, and the 3D error map elements. 內容:3D 頂點資訊區塊包含 3D 頂點座標區塊、3D 頂點識別碼、3D 頂點法線區塊、3D 頂點紋理區塊,以及 3D 錯誤映射元素。
7.65 3D vertex coordinate block 7.65 3D 頂點座標區塊
Abstract values: 3D coordinate Cartesian unsigned short block, see ISO/IEC 39794-1. 抽象值:3D 座標笛卡爾無符號短區塊,請參閱 ISO/IEC 39794-1。
Contents: The location of each vertex is represented by its X coordinate, Y coordinate, and Z coordinate. 內容:每個頂點的位置由其 X 座標、Y 座標和 Z 座標表示。
ISO/IEC 39794-5:2019(E)
7.66 3D vertex identifier 7.66 3D 頂點識別碼
Abstract values: Integer. 抽象值:整數。
Contents: This element shall obtain a unique identifier for the associated vertex. Each two vertices in a record shall have different identifiers. 內容:此元素應取得相關聯頂點的唯一識別碼。記錄中每兩個頂點應具有不同的識別碼。
NOTE If the 3D vertex identifier is absent for a vertex, it is impossible to refer to it in the 3D vertex triangle data block. 注意:如果頂點缺少 3D 頂點識別碼,則無法在 3D 頂點三角形資料區塊中引用該頂點。
7.67 3D vertex normals block 7.67 3D 頂點法線區塊
Abstract values: 3D coordinate Cartesian unsigned short block, see ISO/IEC 39794-1. 抽象值:3D 座標笛卡兒無符號短整數區塊,請參閱 ISO/IEC 39794-1。
Contents: The 3D vertex normals block contains the normal XX, normal YY and normal ZZ coordinate elements. 內容:3D 頂點法線區塊包含法線 XX 、法線 YY 和法線 ZZ 座標元素。
7.68 3D vertex textures block 7.68 3D 頂點紋理區塊
Abstract values: 2D coordinate Cartesian unsigned short block, see ISO/IEC 39794-1. 抽象值:2D 座標笛卡兒無符號短整數區塊,請參閱 ISO/IEC 39794-1。
Contents: The vertex texture X and vertex texture Y fields represent the corresponding x and y pixel position in the 3D texture map block with (0,0)(0,0) denoting the upper left corner. 內容:頂點紋理 X 和頂點紋理 Y 欄位表示 3D 紋理貼圖區塊中對應的 x 和 y 像素位置,其中 (0,0)(0,0) 表示左上角。
7.69 3D error map 7.69 3D 錯誤圖
Abstract values: Octet string. 抽象值:八位元組字串。
Contents: The 3D error map can be used to give further information on how the 3D data has been processed before it was stored in the 3D representation. The 3D error map shall be coded in the PNG format using an 8 bit per pixel greyscale image. The length of the map is variable, as it depends on the lossless compression algorithm. 內容:3D 錯誤圖可用於提供更多資訊,說明 3D 資料在儲存為 3D 表示之前是如何處理的。3D 錯誤圖應以 PNG 格式編碼,使用每像素 8 位元的灰階影像。圖的長度是可變的,因為它取決於無損壓縮演算法。
Pixel values tt in the range of 0 to 199 and 206 to 255 are reserved for future use. A value of t=200t=200 codes that the depth value is considered to be correct. Values of t >= 201t \geq 201 code a specific potential or corrected defect of the 3D data or the corresponding texture image. See Table 6 for an enumerated list of possible values. 像素值 tt 在 0 到 199 以及 206 到 255 的範圍內保留供未來使用。值 t=200t=200 表示深度值被認為是正確的。值 t >= 201t \geq 201 表示 3D 資料或對應紋理影像的特定潛在或已校正缺陷。可能值的列舉清單請參閱表 6。
Table 6-3D error map values 表 6—3D 錯誤圖值
Description 說明
Value 數值
Reserved for future use 保留供未來使用
0 to 199 0 到 199
Depth value is considered correct 深度值被視為正確
200
Depth value is interpolated, interpolation type isn't specified 深度值已內插,但未指定內插類型
201
Depth value is interpolated, linear interpolation has been used 深度值已內插,使用線性內插法
202
Depth value is interpolated, bi-cubic interpolation has been used 深度值已內插,使用雙三次內插法
203
Value of optional texture image potentially wrong (texture noisy, overexposure, etc.) 選用紋理影像的值可能錯誤(紋理雜訊、過度曝光等)
204
Value of optional texture image has been corrected by post processing (image processing) 選用紋理影像的值已透過後處理(影像處理)校正
205
Reserved for future use 保留供未來使用
206 to 255 206 至 255
Description Value
Reserved for future use 0 to 199
Depth value is considered correct 200
Depth value is interpolated, interpolation type isn't specified 201
Depth value is interpolated, linear interpolation has been used 202
Depth value is interpolated, bi-cubic interpolation has been used 203
Value of optional texture image potentially wrong (texture noisy, overexposure, etc.) 204
Value of optional texture image has been corrected by post processing (image processing) 205
Reserved for future use 206 to 255| Description | Value |
| :--- | :--- |
| Reserved for future use | 0 to 199 |
| Depth value is considered correct | 200 |
| Depth value is interpolated, interpolation type isn't specified | 201 |
| Depth value is interpolated, linear interpolation has been used | 202 |
| Depth value is interpolated, bi-cubic interpolation has been used | 203 |
| Value of optional texture image potentially wrong (texture noisy, overexposure, etc.) | 204 |
| Value of optional texture image has been corrected by post processing (image processing) | 205 |
| Reserved for future use | 206 to 255 |
7.70 3D vertex triangle data block 7.70 3D 頂點三角形資料區塊
Abstract values: None. 抽象值:無。
Contents: The 3D vertex triangle data block contains a list of triangle descriptions. Each triangle is specified by the three indices (Triangle index 1, Triangle index 2, and Triangle index 3 ) of the vertices in the vertex data list forming the triangle. The order of the vertex indices shall be counter clock wise to indicate the external face of the triangle. 內容:3D 頂點三角形資料區塊包含三角形描述的列表。每個三角形由構成該三角形的頂點資料列表中的三個頂點索引(三角形索引 1、三角形索引 2 和三角形索引 3)指定。頂點索引的順序應為逆時針方向,以指示三角形的外表面。
7.71 3D coordinate system 7.71 3D 座標系統
Abstract values: 3D Cartesian coordinate system. 抽象值:3D 笛卡兒座標系統。
Contents: This element contains information on the coordinate system used. 內容:此元素包含所用座標系統的資訊。
Originally, 3D data is acquired in a device dependent coordinate system. Based on the knowledge about several device parameters the 3D data can be transformed in Cartesian coordinates. This transformation may involve rotation, translation and resampling. Efforts must be made to preserve the precision of the original data as intended by this document and defined by the 3D textured image resolution block. 最初,3D 資料是在與裝置相關的座標系統中獲取的。根據對多個裝置參數的了解,3D 資料可以轉換為笛卡兒座標。此轉換可能涉及旋轉、平移和重新取樣。必須努力保持原始資料的精確度,如同本文件所預期並由 3D 紋理影像解析度區塊所定義。
This document supports the Cartesian coordinate system for all encodings. 本文件支援所有編碼的笛卡爾座標系統。
The transformation to metric world coordinates is described by appropriate scaling factors and implicit rules (e.g. as used in the anthropometric landmark type). 轉換為公制世界座標的過程,是透過適當的縮放因子和隱含規則來描述(例如,人體測量學地標類型中使用的規則)。
7.72 3D Cartesian coordinate system 7.72 3D 笛卡爾座標系統
Abstract values: None. 抽象值:無。
Contents: In the 3D Cartesian coordinate system, the point of origin must be defined in order to get positive encoding of XYZ coordinates. 內容:在 3D 笛卡爾座標系統中,必須定義原點才能獲得 XYZ 座標的正編碼。
Figure 11 shows two examples of a metric Cartesian coordinate system. In the left, a sample of a Cartesian coordinate system with the origin on the tip of the nose is shown. The XZ plane is defined parallel to the Frankfurt Horizon. In the right, a sample of a metric Cartesian coordinate system with the origin at the middle of the two eyes is given. The XZ plane passes the horizontal gaze axis. This metric Cartesian coordinate system is used by the 3D textured face image application profile. The X axis leads from right eye to left eye, the Z axis is in Horizontal eye direction, looking straight forward in rest position. 圖 11 顯示了度量笛卡爾座標系統的兩個範例。左側顯示了一個以鼻尖為原點的笛卡爾座標系統範例。XZ 平面定義為平行於法蘭克福平面。右側顯示了一個以兩眼之間中點為原點的度量笛卡爾座標系統範例。XZ 平面通過水平凝視軸。此度量笛卡爾座標系統由 3D 紋理臉部影像應用程式設定檔使用。X 軸從右眼指向左眼,Z 軸位於水平眼方向,在靜止位置時直視前方。
Key 圖例
X, Y, Z coordinate axes X、Y、Z 座標軸
FH Frankfurt Horizon FH 法蘭克福地平線
0 coordinate origin 0 座標原點
Figure 11 - Samples of Cartesian coordinate systems 圖 11 - 笛卡兒座標系統範例
7.73 3D Cartesian scales and offsets block 7.73 3D 笛卡兒比例與偏移區塊
Abstract values: Real. 抽象值:實數。
Contents: ScaleX, ScaleY, ScaleZ, OffsetX, OffsetY and OffsetZ are needed to transform digital coordinates to metric coordinates. The scale values have no dimension, the offset values are given in millimetres. 內容:需要 ScaleX、ScaleY、ScaleZ、OffsetX、OffsetY 和 OffsetZ 來將數位座標轉換為公制座標。比例值沒有維度,偏移值以毫米為單位。
The transformation from Cartesian coordinates to metric Cartesian coordinates is derived as follows: 從笛卡爾座標到公制笛卡爾座標的轉換推導如下:
There is a strong relation between anthropometric landmarks and the metric Cartesian coordinate system, as the landmarks define the origin and the orientation. 人體測量學標誌點與公制笛卡爾座標系統之間存在密切關係,因為標誌點定義了原點和方向。
For certain 3D face image kinds, the origin of the metric Cartesian coordinate system can be the midpoint between the left eye centre (12.1) and the right eye centre (12.2), or can be also the nose (prn). 對於某些 3D 臉部影像類型,公制笛卡爾座標系統的原點可以是左眼中心 (12.1) 和右眼中心 (12.2) 之間的中點,也可以是鼻子 (prn)。
For certain 3D face image kinds, the orientation of the Cartesian system is linked to the pose of the head. One example is the frontal pose which is defined by the Frankfurt Horizon as the xz plane and the vertical symmetry plane as the yz plane with the z axis oriented in the direction of the face sight. Another example is the rest position (gaze looking straight forward) with the xzx z plane passing the two eye centres, and the horizontal gaze axis, vertical symmetry plane as the yz plane with the z axis oriented in the direction of the face sight. 對於某些 3D 臉部影像類型,笛卡兒座標系統的方向與頭部姿態相關。其中一個例子是正面姿態,其定義為以法蘭克福平面作為 xz 平面,垂直對稱平面作為 yz 平面,z 軸指向臉部視線方向。另一個例子是靜止位置(凝視正前方),其中 xzx z 平面穿過兩個眼睛中心,水平凝視軸、垂直對稱平面作為 yz 平面,z 軸指向臉部視線方向。
Large values of ScaleX, ScaleY or ScaleZ indicate a low spatial sampling rate in the respective dimension. Boundary values of ScaleX, ScaleY and ScaleZ may be strongly restricted for different 3D face image kinds. ScaleX、ScaleY 或 ScaleZ 的值較大表示在各自維度上的空間取樣率較低。ScaleX、ScaleY 和 ScaleZ 的邊界值可能會因不同的 3D 臉部影像類型而受到嚴格限制。
7.74 3D face image kind 7.74 3D 臉部影像類型
Abstract values: None. 抽象值:無。
Contents: The 3D face image kind element shall represent the type of the face image stored in the 3D representation data. See Table 7 for a list of allowed image types and their normative requirements. 內容:3D 臉部影像類型元素應表示儲存在 3D 表示資料中的臉部影像類型。允許的影像類型及其規範要求請參閱表 7。
Table 7 - 3D Face image kind codes 表 7 - 3D 臉部影像種類代碼
Value 數值
Definition and normative requirements 定義與規範要求
3D Textured face images 3D 紋理臉部影像
Annex D.3 附錄 D.3
Value Definition and normative requirements
3D Textured face images Annex D.3| Value | Definition and normative requirements |
| :---: | :--- |
| 3D Textured face images | Annex D.3 |
7.75 3D physical face measurements block 7.75 3D 實體臉部測量區塊
Abstract values: None. 抽象值:無。
Contents: For specific application domains different minimal spatial sampling rates of the interchange data may be required. For example, using higher spatial sampling rate images allow for specific human as well as machine inspection methods that depend on the analysis of very small details. 內容:對於特定的應用領域,可能需要不同的最小空間取樣率交換資料。例如,使用較高空間取樣率的影像,可支援依賴於分析極小細節的特定人工及機器檢測方法。
The 3D physical face measurements block consists of four elements. If the width of the head shall be encoded the 3D physical head width may be used. If the length of the head shall be encoded the 3D physical head length may be used. If the inter-eye distance shall be encoded the Physical inter-eye distance may be used. If the eye-tomouth distance shall be encoded the 3D physical eye-to-mouth distance may be used. If necessary, all four elements may be used. All measures shall be given in millimetres. See 7.48 for equivalent definitions for pixel measurements in 2D images. 3D 實體臉部測量區塊包含四個元素。若要編碼頭部寬度,可使用 3D 實體頭部寬度。若要編碼頭部長度,可使用 3D 實體頭部長度。若要編碼眼距,可使用實體眼距。若要編碼眼到嘴距離,可使用 3D 實體眼到嘴距離。如有必要,可使用所有四個元素。所有測量值應以毫米為單位。2D 影像中像素測量的等效定義請參閱 7.48。
7.76 3D physical head width 7.76 3D 實體頭部寬度
Abstract values: Integer. 抽象值:整數。
Contents: The 3D physical head width element provides information on the width of the head in millimetres. 內容:3D 實體頭部寬度元素提供以毫米為單位的頭部寬度資訊。
7.77 3D physical inter-eye distance 7.77 3D 實體眼距
Abstract values: Integer. 抽象值:整數。
Contents: The 3D physical inter-eye distance element provides information on the distance between the eye midpoints in millimetres. 內容:3D 實體眼距元素提供以公釐為單位的兩眼中心點間距離資訊。
7.78 3D physical eye-to-mouth distance 7.78 3D 實體眼口距離
Abstract values: Integer. 抽象值:整數。
Contents: The 3D physical eye-to-mouth distance element provides information on the distance between the mouth and the eyes in millimetres, more precise, between the midpoint of the line connecting the eye centres (feature points 12.1 and 12.2) and the mouth (feature point 2.3). 內容:3D 實體眼口距離元素提供以公釐為單位的嘴巴與眼睛間距離資訊,更精確地說,是連接兩眼中心點(特徵點 12.1 和 12.2)的線段中點與嘴巴(特徵點 2.3)之間的距離。
7.79 3D physical head length 7.79 3D 實體頭部長度
Abstract values: Integer. 抽象值:整數。
Contents: The 3D physical head length element provides information on the distance from the chin to crown, or length, of the head, in millimetres. 內容:3D 實體頭部長度元素提供從下巴到頭頂的距離,或頭部以毫米為單位的長度資訊。
7.80 3D textured image resolution block 7.80 3D 紋理影像解析度區塊
Abstract values: None. 抽象值:無。
Contents: The 3D textured image resolution block consists of MM shape [X/Y/Z] resolution, 3D MM texture resolution, 3D texture acquisition period, and 3D face area scanned. 內容:3D 紋理影像解析度區塊包含 MM 形狀 [X/Y/Z] 解析度、3D MM 紋理解析度、3D 紋理擷取週期,以及掃描的 3D 臉部區域。
7.81 3D MM shape [X/Y/Z] resolution 7.81 3D MM 形狀 [X/Y/Z] 解析度
Abstract values: Real (Decimal). 抽象值:實數(十進位)。
Contents: The 3D MM shape X resolution, the 3D MM shape Y resolution, and the 3D MM shape Z resolution define the minimal distance acquired by the shape acquisition system in millimetres. These resolutions may be different compared with the MM texture resolution value. 內容:3D MM 形狀 X 解析度、3D MM 形狀 Y 解析度,以及 3D MM 形狀 Z 解析度,定義了形狀擷取系統以毫米為單位擷取的最小距離。這些解析度可能與 MM 紋理解析度值不同。
7.82 3D MM texture resolution 7.82 3D MM 紋理解析度
Abstract values: Real (Decimal). 抽象值:實數(十進位)。
Contents: The 3D MM texture resolution defines the minimal distance acquired by the texture acquisition system in mm. This resolution may be different compared with the 3D MM shape [X/Y/Z] resolution values. 內容:3D MM 紋理解析度定義了紋理擷取系統所擷取的最小距離(單位為毫米)。此解析度可能與 3D MM 形狀 [X/Y/Z] 解析度值不同。
7.83 3D texture acquisition period 7.83 3D 紋理擷取週期
Abstract values: Real (Decimal). 抽象值:實數(十進位)。
Contents: The 3D texture acquisition period defines the time in milliseconds used for shape and texture acquisition. During this period neither the acquisition system nor the subject shall move or be moved. 內容:3D 紋理擷取週期定義了用於形狀和紋理擷取的時間(以毫秒為單位)。在此期間,擷取系統和受測者均不得移動或被移動。
7.84 3D face area scanned block 7.84 3D 臉部區域掃描區塊
Abstract values: The value of this element shall be one or more of the following: 抽象值:此元素的值應為以下其中一個或多個:
Front of the head; 頭部正面;
Chin; 下巴;
Ears; 耳朵;
Neck; 頸部;
Back of the head; 後腦勺;
Full head. 整個頭部。
Contents: The 3D face area scanned shall indicate the area scanned of the face. The minimum allowed 3D face area scanned is Front of the head. 內容:掃描的 3D 臉部區域應指示臉部掃描的區域。允許的最小 3D 臉部掃描區域是頭部正面。
7.85 3D texture map block 7.85 3D 紋理貼圖區塊
Abstract values: None. 抽象值:無。
Contents: The 3D texture map block consists of the 3D texture map data, the Image data format, the 3D texture capture device spectral block, the 3D texture standard illuminant, and the 3D error map (See 7.70) elements. 內容:3D 紋理圖區塊包含 3D 紋理圖資料、影像資料格式、3D 紋理擷取裝置光譜區塊、3D 紋理標準光源,以及 3D 錯誤圖(參見 7.70)等元素。
The 3D texture map block should only be used to store face texture data that is acquired by a scanning device during the 3D acquisition process, and therefore may have a different geometry than the 2D representation data stored in the same BDB. It is not a substitute for the 2D representation data. The 3D texture map shall be coded in 8 bit or 16 bit greyscale or as a 24 bit colour image. The length of the map is variable as it depends on the applied compression algorithm. 3D 紋理圖區塊僅應用於儲存透過掃描裝置在 3D 擷取過程中取得的臉部紋理資料,因此其幾何形狀可能與儲存在相同 BDB 中的 2D 表示資料不同。它不能替代 2D 表示資料。3D 紋理圖應以 8 位元或 16 位元灰階編碼,或以 24 位元彩色影像編碼。地圖的長度是可變的,因為它取決於所應用的壓縮演算法。
7.86 3D texture capture device spectral block 7.86 3D 紋理擷取裝置光譜區塊
Abstract values: The value of this element shall be one of the following: 摘要值:此元素的數值應為下列其中之一:
Contents: The 3D texture capture device spectral block denotes the kind of spectrum that has been used for acquiring the 3D texture map. This spectrum may differ from the one used for 2D image representation data. 內容:3D 紋理擷取裝置光譜區塊表示用於擷取 3D 紋理圖的光譜類型。此光譜可能與用於 2D 影像表示資料的光譜不同。
ISO/IEC 39794-5:2019(E)
7.87 3D texture standard illuminant 7.87 3D 紋理標準光源
Abstract values: The value of this element shall be one of the following: 摘要值:此元素的數值應為下列其中之一:
D30; D30;
D35; D35;
D40; D40;
D45; D45;
D50; D50;
D55; D55;
D60; D60;
D65; D65;
D70; D70;
D75; D75;
D80. D80。
Contents: Illumination according to one of the standard illuminants defined in ISO 11664-2 or similar. 內容:根據 ISO 11664-2 或類似標準中定義的其中一種標準光源進行照明。
7.88 3D texture map data 7.88 3D 紋理貼圖資料
Abstract values: Octet string. 抽象值:八位元字串。
Contents: The 3D texture map data shall contain the face texture data acquired by a capture device during the 3D acquisition process. The 3D texture map data element shall have the format specified in the Image data format element. 內容:3D 紋理貼圖資料應包含擷取裝置在 3D 擷取過程中取得的臉部紋理資料。3D 紋理貼圖資料元素應具有影像資料格式元素中指定的格式。
8 Encoding 8 編碼
8.1 Overview 8.1 概述
The tagged binary encoding as well as the XML encoding is given in this clause and Annex A, respectively. In order to aid recognition of abstract values, the same lower camel-case notation is used for abstract data elements in the ASN. 1 module and in the XSD. The lower camel-case names are derived from the abstract values given here. 帶標籤的二進位編碼和 XML 編碼分別在本條款和附錄 A 中給出。為了幫助識別抽象值,ASN.1 模組和 XSD 中的抽象資料元素均使用相同的駝峰式命名法(lower camel-case)。駝峰式命名法(lower camel-case)的名稱源自此處給出的抽象值。
The names of the ASN. 1 module and of the XML schema definition (available at http://standards.iso org/iso-iec/39794/-5/ed-1/en) are iso-iec-39794-5-ed-1-v1.asn and iso-iec-39794-5-ed-1-v1.xsd, respectively. ASN.1 模組和 XML 綱要定義(可於 http://standards.iso.org/iso-iec/39794/-5/ed-1/en 取得)的名稱分別為 iso-iec-39794-5-ed-1-v1.asn 和 iso-iec-39794-5-ed-1-v1.xsd。
Content and semantics of parameters of ISO/IEC 19794-5 (2011 edition) served as starting point for this document. The syntax has been modified to accommodate new requirements, and many parameters have been added allowing the encoding of many more properties of face images than before. ISO/IEC 19794-5 (2011 年版) 參數的內容和語義是本文件的起點。語法已修改以適應新要求,並增加了許多參數,允許編碼比以前更多的臉部影像屬性。
Most of the face image data record parameters are considered as optional to allow application specific profiles and efficient storage of the available data. 大多數臉部影像資料記錄參數都被視為可選,以允許應用程式特定的設定檔和有效率地儲存可用資料。
The 3D encoding types 3D point map and range image are not supported by this version of this document. 本文件的此版本不支援 3D 編碼類型中的 3D 點陣圖和距離影像。
8.2 Tagged binary encoding 8.2 標記二進位編碼
This clause specifies the ASN. 1 module implementing the abstract data elements specified in Clause 7. It describes the parameters of face image data as they are encoded in ASN.1. These ASN. 1 definitions are based on the following design decisions: 本條款規定了實作第 7 條中指定之抽象資料元素的 ASN.1 模組。它描述了臉部影像資料以 ASN.1 編碼時的參數。這些 ASN.1 定義基於以下設計決策:
The ASN. 1 types as defined in Clause A. 1 which encode the abstract data elements of Clause 7 shall conform to the ASN. 1 standard (ISO/IEC 8824-1) and to ISO/IEC 39794-1. 在條款 A.1 中定義的 ASN.1 類型,其編碼條款 7 的抽象資料元素時,應符合 ASN.1 標準 (ISO/IEC 8824-1) 和 ISO/IEC 39794-1。
The tagged binary encoding of face image data shall be obtained by applying the ASN. 1 distinguished encoding rules (DER) defined in ISO/IEC 8825-1 to a value of the type FaceImageDataBlock defined in the given ASN. 1 module. The DER encoding of each data object has three parts: tag octets that identify the data object, length octets that give the number of subsequent value octets, and the value octets. 臉部影像資料的標記二進位編碼,應透過將 ISO/IEC 8825-1 中定義的 ASN.1 區別編碼規則 (DER) 應用於給定 ASN.1 模組中定義的 FaceImageDataBlock 類型值來取得。每個資料物件的 DER 編碼包含三個部分:識別資料物件的標籤八位元組、指示後續值八位元組數量的長度八位元組,以及值八位元組。
The ISO/IEC 39794 ASN. 1 modules are defined independently, i.e. no re-use or imports of definitions outside the ISO/IEC 39794 series area in order to avoid interdependencies to other standardization bodies even if this might be useful (e.g., considering X.509/PKIX definitions). ISO/IEC 39794 ASN.1 模組是獨立定義的,亦即,為了避免與其他標準化機構產生相互依賴性,即使這可能很有用(例如,考慮 X.509/PKIX 定義),也不會重複使用或匯入 ISO/IEC 39794 系列範圍之外的定義。
Any face image data specific definition is fully included in the ASN. 1 module in this document, any re-usable header field that is defined in the ISO/IEC 39794-1 framework is part of the separate ISO/IEC 39794-1 ASN. 1 module. 任何臉部影像資料的特定定義都完全包含在本文件中的 ASN.1 模組中,而在 ISO/IEC 39794-1 框架中定義的任何可重複使用的標頭欄位,則是獨立的 ISO/IEC 39794-1 ASN.1 模組的一部分。
The entry point for any ISO/IEC 39794 series biometric type definition is the BiometricDataBlock defined in the ISO/IEC 39794-1 ASN. 1 module. This module includes the ASN. 1 definition of all modality specific parts of the ISO/IEC 39794 series. This allows modifying or extending both the generic header information and the supported set of biometric data types at one place without impacting the other parts of the ISO/IEC 39794 series. For example, the ISO/IEC 39794-1 ASN. 1 module includes the definitions of face image data and fingerprint image data and is extended later on by iris data. In this case, the ASN. 1 definitions of ISO/IEC 39794-4 and this document do not need to be modified. 任何 ISO/IEC 39794 系列生物特徵類型定義的進入點,都是 ISO/IEC 39794-1 ASN.1 模組中定義的 BiometricDataBlock。此模組包含 ISO/IEC 39794 系列所有模態特定部分的 ASN.1 定義。這允許在一個地方修改或擴充通用標頭資訊和支援的生物特徵資料類型集,而不會影響 ISO/IEC 39794 系列的其他部分。例如,ISO/IEC 39794-1 ASN.1 模組包含臉部影像資料和指紋影像資料的定義,並在之後透過虹膜資料進行擴充。在這種情況下,ISO/IEC 39794-4 和本文件的 ASN.1 定義無需修改。
Extension markers are included in all data elements to ensure extensibility and forward/backward compatibility when new parameters need to be added to existing containers/blocks. 所有資料元素中都包含擴充標記,以確保當需要向現有容器/區塊添加新參數時,具有可擴充性以及向前/向後相容性。
The latest version of the ASN. 1 standard is employed, namely ISO/IEC 8824-1:2015. 採用了最新版本的 ASN.1 標準,即 ISO/IEC 8824-1:2015。
The distinguished encoding rules (DER) as specified in ISO/IEC 8825-1 are utilized to represent the data in binary format. Other options such as XML encoding rules shall not be used. The syntax of face image XML documents shall be based on the XML schema definition in A.2, not on the ASN. 1 module in A.1. 根據 ISO/IEC 8825-1 中規定的區別編碼規則(DER),將資料以二進位格式表示。不應使用其他選項,例如 XML 編碼規則。臉部影像 XML 文件的語法應基於 A.2 中的 XML 綱要定義,而非 A.1 中的 ASN.1 模組。
The ASN. 1 module in A. 1 is available at http://standards.iso.org/iso-iec/39794/-5/ed-1/en. A.1 中的 ASN.1 模組可於 http://standards.iso.org/iso-iec/39794/-5/ed-1/en 取得。
Additional explanations on the mapping between the specifications in Clause 7 and the ASN. 1 module given in A. 1 apply: 關於第 7 條中的規範與 A.1 中給定的 ASN.1 模組之間的對應關係,適用以下額外說明:
The ASN. 1 schema does not guarantee that if all elements that could be contained in an element are absent, the whole element is absent too. ASN.1 綱要不保證如果一個元素中所有可能包含的元素都不存在,則整個元素也不存在。
If in the propertiesBlock element a property is set to TRUE, the respective property is present in the image. Otherwise if its set to FALSE, that property is absent in the image. If a property is omitted no statement has been made. 如果在 propertiesBlock 元素中,某個屬性設定為 TRUE,則該屬性會出現在影像中。否則,如果設定為 FALSE,則該屬性不會出現在影像中。如果某個屬性被省略,則表示沒有做出任何聲明。
If in the expressionBlock element one of the components is set to TRUE, the respective attribute is present in the image. Otherwise if its set to FALSE, it is absent in the image. If an element is omitted no statement has been made. The ASN. 1 schema does not prevent from choosing the expressions neutral and smile for the same face image. However, neutral and smile shall not be both true for the same image. 如果在 expressionBlock 元素中,其中一個元件設定為 TRUE,則該屬性會出現在影像中。否則,如果設定為 FALSE,則該屬性不會出現在影像中。如果省略某個元素,則表示未做出任何陳述。ASN.1 綱要並未阻止為同一張臉部影像選擇「中性」和「微笑」表情。然而,「中性」和「微笑」不應同時為同一張影像的真實狀態。
At least one of the elements of the poseAngleBlock element shall be present; otherwise the whole poseAngleBlock element shall be absent. This requirement is not covered by the ASN. 1 schema. poseAngleBlock 元素中至少應存在一個元素;否則,整個 poseAngleBlock 元素應不存在。此要求未涵蓋在 ASN.1 綱要中。
ISO/IEC 39794-5:2019(E)
MPEG4 feature points with the abstract name < 1 > . < 2 ><1>.<2> are encoded as mpeg4PointCode-<01>-<02>. AnthropometricLandmarkPointCode elements with the abstract name (:1:).(:2:)\langle 1\rangle .\langle 2\rangle are encoded as pointCode-<01>-<02>. 抽象名稱為 < 1 > . < 2 ><1>.<2> 的 MPEG4 特徵點編碼為 mpeg4PointCode-<01>-<02>。抽象名稱為 (:1:).(:2:)\langle 1\rangle .\langle 2\rangle 的 AnthropometricLandmarkPointCode 元素編碼為 pointCode-<01>-<02>。
Encoding examples are contained in Annex B. 編碼範例包含在附錄 B 中。
8.3 XML encoding 8.3 XML 編碼
Annex A. 2 specifies an XSD schema, in which the abstract data elements of Clause 7 are constrained by XML types defined within one of the following standards: W3C Recommendations, XML Schema Parts 1 and 2, ISO/IEC 39794-1, or this document. 附錄 A.2 規定了一個 XSD 綱要,其中第 7 條款的抽象資料元素受到以下標準之一中定義的 XML 類型約束:W3C 建議書、XML 綱要第 1 和第 2 部分、ISO/IEC 39794-1 或本文件。
Binary data shall only be encoded as base 64 and stored as a text string in an element, which itself has the underlying type of xs:base64Binary, for example: <xs:element name=“data” type=“xs:base64Binary”/> 二進位資料應僅以 base 64 編碼,並以文字字串形式儲存在元素中,該元素本身具有 xs:base64Binary 的基礎類型,例如:<xs:element name=“data” type=“xs:base64Binary”/>
For avoidance of doubt other methods of encoding binary data such as xs:hexBinary or proprietary extensions which support binary data encoding (i. e.: XOP) are not permitted. 為避免疑慮,不允許使用其他二進位資料編碼方法,例如 xs:hexBinary 或支援二進位資料編碼的專有擴充功能(即:XOP)。
Additional explanations on the mapping between the specifications in Clause 7 and the XSD given in A. 2 apply: 關於條款 7 中規範與 A.2 中給出的 XSD 之間映射的額外解釋適用:
The XML schema does not guarantee that if all elements that could be contained in an element are absent, the whole element is absent, too. XML 結構描述無法保證,如果某個元素中可能包含的所有元素都不存在,則該整個元素也不存在。
If a property in a propertiesBlock element is set to TRUE, this property is present in the image. Otherwise if its set to FALSE, the property is absent in the image. If a property is omitted no statement has been made. 如果 propertiesBlock 元素中的某個屬性設定為 TRUE,則此屬性存在於影像中。否則,如果設定為 FALSE,則該屬性不存在於影像中。如果省略某個屬性,則表示未做出任何聲明。
If an expression in an expressionBlock element is set to TRUE, this expression is present in the image. Otherwise if its set to FALSE, the expression is absent in the image. If an expression is omitted no statement has been made. 如果 expressionBlock 元素中的某個表達式設定為 TRUE,則此表達式存在於影像中。否則,如果設定為 FALSE,則該表達式不存在於影像中。如果省略某個表達式,則表示未做出任何聲明。
The XML schema does not prevent from choosing the expressions neutral and smile for the same face image. However, neutral and smile shall not be both true for the same image. XML 結構描述並未阻止針對同一個臉部影像選擇「中性」和「微笑」這兩種表達式。然而,對於同一個影像,「中性」和「微笑」不應同時為真。
At least one of the elements of the poseAngleBlock element shall be present; otherwise the whole poseAngleBlock element shall be absent. This requirement is not covered by the XML schema. poseAngleBlock 元素中至少應存在一個元素;否則,整個 poseAngleBlock 元素應不存在。此要求未涵蓋在 XML 綱要中。
MPEG4 feature points with the abstract name < 1 > . < 2 ><1>.<2> are encoded as MPEG4PointCode-<01>-<02>. 抽象名稱為 < 1 > . < 2 ><1>.<2> 的 MPEG4 特徵點編碼為 MPEG4PointCode-<01>-<02>。
AnthropometricLandmarkPointCode elements with the abstract name (:1:).(:2:)\langle 1\rangle .\langle 2\rangle are encoded as PointCode-<01>-<02>. 抽象名稱為 (:1:).(:2:)\langle 1\rangle .\langle 2\rangle 的 AnthropometricLandmarkPointCode 元素編碼為 PointCode-<01>-<02>。
The XSD module in A. 2 can be retrieved from http://standards.iso.org/iso-iec/39794/-5/ed-1/en. A.2 中的 XSD 模組可從 http://standards.iso.org/iso-iec/39794/-5/ed-1/en 取得。
Encoding examples are contained in Annex B. 編碼範例請參閱附錄 B。
9 Registered BDB format identifiers 9 已註冊的 BDB 格式識別碼
The registrations listed in Table 8 have been made in accordance with ISO/IEC 19785 (all parts) ^([31]){ }^{[31]} to identify the face image data interchange formats defined in this document. The format owner is ISO/ IEC JTC 1/SC 37 with the registered biometric organization identifier 257 (0101Hex). 表 8 所列的註冊是依據 ISO/IEC 19785 (所有部分) ^([31]){ }^{[31]} 進行的,旨在識別本文件所定義的臉部影像資料交換格式。格式擁有者為 ISO/IEC JTC 1/SC 37,其註冊的生物識別組織識別碼為 257 (0101Hex)。
ISO-IEC-39794-5-ed-1-v1 {iso(1) standard(0) iso-iec-39794(39794) part-5(5) ed-1(1) v1(1)
iso-iec-39794-5(0) }
-- Use of ISO/IEC copyright in this Schema is licensed for the purpose of
-- developing, implementing, and using software based on this Schema, subject
-- to the following conditions:
--
-- * Software developed from this Schema must retain the Copyright Notice,
-- this list of conditions and the disclaimer below ("Disclaimer").
--
-- * Neither the name or logo of ISO or of IEC, nor the names of specific
-- contributors, may be used to endorse or promote software derived from
-- this Schema without specific prior written permission.
--
-- * The software developer shall attribute the Schema to ISO/IEC and
-- identify the ISO/IEC standard from which it is taken. Such attribution
-- (e.g., "This software makes use of the Schema from ISO/IEC 39794-5
-- within modifications permitted in the relevant ISO/IEC standard.
-- Please reproduce this note if possible."), may be placed in the
-- software itself or any other reasonable location.
-- The Disclaimer is:
-- THE SCHEMA ON WHICH THIS SOFTWARE IS BASED IS PROVIDED BY THE COPYRIGHT
-- HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES,
-- INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY
-- AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL
-- THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
-- INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
-- NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
-- DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
-- THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-- (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
-- THE CODE COMPONENTS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
DEFINITIONS IMPLICIT TAGS ::= BEGIN
IMPORTS
VersionBlock,
CaptureDateTimeBlock,
QualityBlocks,
PADDataBlock,
CoordinateCartesian2DUnsignedShortBlock,
CoordinateCartesian3DUnsignedShortBlock,
RegistryIdBlock,
CertificationIdBlocks
FROM ISO-IEC-39794-1-ed-1-v1;
FaceImageDataBlock ::= [APPLICATION 5] SEQUENCE {
versionBlock [0] VersionBlock,
representationBlocks [1] RepresentationBlocks,
}
<?xml version="1.0" encoding="utf-8"?>
<!--Use of ISO/IEC copyright in this Schema is licensed for the purpose of developing,
implementing, and using software based on this Schema, subject to the following
conditions:
* Software developed from this Schema must retain the Copyright Notice, this list of
conditions and the disclaimer below ("Disclaimer").
ISO/IEC 39794-5:2019(E)
Neither the name or logo of ISO or of IEC, nor the names of specific contributors, may be used to endorse or promote software derived from this Schema without specific prior written permission. 未經特定書面許可,不得使用 ISO 或 IEC 的名稱或標誌,或特定貢獻者的名稱,來認可或推廣源自本綱要的軟體。
The software developer shall attribute the Schema to ISO/IEC and identify the ISO/IEC standard from which it is taken. Such attribution (e.g., “This software makes use of the Schema from ISO/IEC 39794-5 within modifications permitted in the relevant ISO/IEC standard. Please reproduce this note if possible.”), may be placed in the software itself or any other reasonable location. 軟體開發者應將此綱要歸因於 ISO/IEC,並標明其所取用的 ISO/IEC 標準。此類歸因(例如:「本軟體使用來自 ISO/IEC 39794-5 的綱要,並在相關 ISO/IEC 標準允許的範圍內進行修改。如可能,請轉載此註記。」)可放置於軟體本身或任何其他合理位置。
The Disclaimer is: 免責聲明為:
THE SCHEMA ON WHICH THIS SOFTWARE IS BASED IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS “AS IS” AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THE CODE COMPONENTS, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.–> 本軟體所依據的架構係由著作權人及貢獻者「按現狀」提供,不提供任何明示或暗示之擔保,包括但不限於適售性及特定用途適用性之暗示擔保。在任何情況下,著作權人或貢獻者均不對任何直接、間接、偶發、特殊、懲戒性或衍生性損害(包括但不限於採購替代商品或服務;使用、資料或利潤之損失;或業務中斷)負責,無論其發生原因及責任理論為何,亦無論係因契約、嚴格責任或侵權行為(包括過失或其他)所致,即使已告知可能發生此類損害,亦同。–>
<xs:schema
xmlns:xs=“http://www.w3.org/2001/XMLSchema”
xmlns:vc=“http://www.w3.org/2007/XMLSchema-versioning”
xmlns:cmn=“http://standards.iso.org/iso-iec/39794/-1”
xmlns=“http://standards.iso.org/iso-iec/39794/-5”
vc:minVersion=“1.0”
targetNamespace=“http://standards.iso.org/iso-iec/39794/-5”
elementFormDefault=“qualified”
attributeFormDefault=“unqualified”>
<xs:import namespace=“http://standards.iso.org/iso-iec/39794/-1” schemaLocation=“iso-iec-39794-1-ed-1-v1.xsd” />
<xs:element name=“faceImageData” type=“FaceImageDataBlockType”> xs:annotation xs:documentationroot element</xs:documentation>
</xs:annotation>
</xs:element>
<xs:complexType name=“FaceImageDataBlockType”> xs:sequence
<xs:element name=“versionBlock” type=“cmn:VersionBlockType” />
<xs:element name=“representationBlocks” type=“RepresentationBlocksType” />
<xs:any namespace="##other" processContents=“lax” minOccurs=“0” />
</xs:sequence>
</xs:complexType>
<xs:complexType name=“RepresentationBlocksType”> xs:sequence
<xs:element name=“representationBlock” type=“RepresentationBlockType”
maxOccurs=“unbounded” />
</xs:sequence>
</xs:complexType>
<xs:complexType name=“RepresentationBlockType”> xs:sequence
<xs:element name=“representationId” type=“xs:unsignedInt” />
<xs:element name=“imageRepresentation” type=“ImageRepresentationType” />
<xs:element name=“captureDateTimeBlock” type=“cmn:CaptureDateTimeBlockType”
minOccurs=“0” />
<xs:element name=“qualityBlocks” type=“cmn:QualityBlocksType” minOccurs=“0” />
<xs:element name=“padDataBlock” type=“cmn:PADDataBlockType” minOccurs=“0” />
<xs:element name=“sessionId” type=“xs:unsignedInt” minOccurs=“0” />
<xs:element name=“derivedFrom” type=“xs:unsignedInt” minOccurs=“0” />
<xs:element name=“captureDeviceBlock” type=“CaptureDeviceBlockType” minOccurs=“0” /> <xs:element name=“identityMetadataBlock” type=“IdentityMetadataBlockType”
minOccurs=“0” />
<xs:element name=“landmarkBlocks” type=“LandmarkBlocksType” minOccurs=“0” />
A binary TLV encoding example based on the ASN. 1 schema in Annex A. 1 is given below. This example encoding is available at http://standards.iso.org/iso-iec/39794/-5/ed-1/en. 以下提供一個基於附錄 A.1 中 ASN.1 綱要的二進位 TLV 編碼範例。此編碼範例可於 http://standards.iso.org/iso-iec/39794/-5/ed-1/en 取得。
B. 2 Example encodings for a face image XML document B. 2 臉部影像 XML 文件的編碼範例
An encoding example based on the XSD schema in Annex A. 2 is given below. This example encoding is available at http://standards.iso.org/iso-iec/39794/-5/ed-1/en. 以下是根據附錄 A. 2 中 XSD 綱要的編碼範例。此編碼範例可於 http://standards.iso.org/iso-iec/39794/-5/ed-1/en 取得。
NOTE The example available at http://standards.iso.org/iso-iec/39794/-5/ed-1/en possesses full contents of the 2D face JPEG image data encoded within the base64Binary type value field of the “representationData2D” element. The value is replaced below in a truncated version (with full length in the brackets) for simplicity. 註:http://standards.iso.org/iso-iec/39794/-5/ed-1/en 所提供的範例,其「representationData2D」元素的 base64Binary 類型數值欄位中,包含完整的 2D 臉部 JPEG 影像資料。為求簡潔,以下將該數值替換為截斷版本(括號內為完整長度)。
This annex specifies elements of the conformance testing methodology, test assertions, and test procedures as applicable to this document. Specifically it establishes: 本附錄規定了適用於本文件的符合性測試方法、測試斷言和測試程序的要素。具體而言,它確立了:
test assertions of the structure of the face image data format as specified in this document (Type A Level 1), 測試本文件規定之臉部影像資料格式結構的斷言(A 型等級 1),
test assertions of internal consistency by checking the types of values that may be contained within each element (Type A Level 2), 透過檢查每個元素中可能包含的值類型來測試內部一致性(A 型等級 2),
tests of semantic assertions (Type A Level 3). 語義斷言測試(A 型等級 3)。
This conformance testing methodology does not establish: 本一致性測試方法並未確立:
tests of conformance of CBEFF structures required by ISO/IEC 39794-1, 符合 ISO/IEC 39794-1 要求的 CBEFF 結構一致性測試,
tests of conformance of the image data to the quality-related specifications, 影像資料與品質相關規範的一致性測試,
tests of conformance of the image data blocks to the respective JPEG or JPEG 2000 standards, 影像資料區塊與各自 JPEG 或 JPEG 2000 標準的一致性測試,
tests of other characteristics of biometric products or other types of testing of biometric products (e.g., acceptance, performance, robustness, security). 生物辨識產品其他特性測試或生物辨識產品其他類型測試(例如,驗收、效能、穩固性、安全)。
To provide sufficient information about the IUT for the testing laboratory to properly conduct a conformance test and for an appropriate declaration of conformity to be made, the supplier of the IUT shall provide the identification of the supplier and the IUT information in Table C. 1 and also complete the columns IUT support and supported range in Table C. 2 that applies to tested face image extensible BDB format(s). All tables shall be provided to the testing laboratory prior to, or at the same time as, the IUT is provided to the testing laboratory. 為使測試實驗室能正確執行一致性測試,並做出適當的一致性聲明,IUT 供應商應提供表 C.1 中供應商和 IUT 的識別資訊,並填寫表 C.2 中適用於已測試臉部影像可擴展 BDB 格式的 IUT 支援和支援範圍欄位。所有表格應在 IUT 提供給測試實驗室之前或同時提供給測試實驗室。
NOTE W3C maintains a list of tools that can be used to work with xml documents and schemas [35]. ITU-T maintains a list of tools that can be used to work with ASN. 1 documents and schemas ^([36]){ }^{[36]}. Validating documents with the schemas will assure all Level 1 conformance issues. 註:W3C 維護一份可用於處理 XML 文件和綱要的工具清單 [35]。ITU-T 維護一份可用於處理 ASN.1 文件和綱要的工具清單 ^([36]){ }^{[36]} 。使用綱要驗證文件將確保所有 Level 1 的一致性問題。
Table C. 1 - Identification of the supplier and the IUT 表 C.1 - 供應商和 IUT 的識別資訊
Supplier name and address 供應商名稱和地址
Contact point for queries about the ICS ICS 查詢聯絡方式
Implementation name 實作名稱
Implementation version 實作版本
Any other information necessary for full identification of the implementation 完整識別實作所需的任何其他資訊
Registered BDB format identifier of the format that conformance is claimed to 註冊的 BDB 格式識別碼,聲明符合該格式
Are any mandatory requirements of the standard not fully supported (Yes or No) 標準的任何強制性要求是否未完全支援(是或否)
Date of statement 聲明日期
Supplier name and address
Contact point for queries about the ICS
Implementation name
Implementation version
Any other information necessary for full identification of the implementation
Registered BDB format identifier of the format that conformance is claimed to
Are any mandatory requirements of the standard not fully supported (Yes or No)
Date of statement | Supplier name and address | |
| :--- | :--- |
| Contact point for queries about the ICS | |
| Implementation name | |
| Implementation version | |
| Any other information necessary for full identification of the implementation | |
| Registered BDB format identifier of the format that conformance is claimed to | |
| Are any mandatory requirements of the standard not fully supported (Yes or No) | |
| Date of statement | |
C. 2 Requirements and options C. 2 要求與選項
Table C. 2 lists the syntactic options and semantic conformance requirements specified in this document. The supplier of the IUT can explain which optional components are supported and the testing laboratory can note the results of the test. Support is defined as the ability of the used structure to fulfil the requirements automatically without further testing. Support does not mean that the requirement can’t be fulfilled when using the structure, all requirements in this table can be fulfilled for both ASN. 1 and XML. 表 C. 2 列出了本文件所規定的語法選項和語義一致性要求。IUT 的供應商可以說明支援哪些可選元件,而測試實驗室可以記錄測試結果。支援的定義是所使用的結構能夠自動滿足要求,無需進一步測試。支援並不表示在使用該結構時無法滿足要求,此表中的所有要求對於 ASN.1 和 XML 都可以滿足。
Table C. 2 details the Level 2 conformance tests that a testing organization should perform on an IUT. These Level 2 tests are necessary as the schema validation does not perform those checks. All other Level 1 and Level 2 conformance requirements are tested by schema validation. 表 C. 2 詳細說明了測試機構應對 IUT 執行的二級一致性測試。這些二級測試是必要的,因為綱要驗證不執行這些檢查。所有其他一級和二級一致性要求均透過綱要驗證進行測試。
Table C. 2 - Requirements and options of the data format specification 表 C. 2 - 資料格式規範的要求和選項
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P1
Annex A 附錄 A
A face-image data block may contain unknown extensions. 人臉影像資料區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P2
Annex A 附錄 A
A representation block may contain a capture date/time block. 一個表示區塊可能包含一個擷取日期/時間區塊。
1 and 2 1 和 2
0
Y
Y
P3
AnnexA 附錄 A
A representation block may contain quality blocks. 一個表示區塊可能包含品質區塊。
1 and 2 1 和 2
0
Y
Y
P4
ISO/IEC 39794-1
A quality block may contain unknown extensions. 品質區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P5
Annex A 附錄 A
A representation block may contain a PAD data block. 一個表示區塊可能包含一個 PAD 資料區塊。
1 and 2 1 和 2
0
Y
Y
P6
ISO/IEC 39794-1
A PAD data block may contain a PAD decision. 一個 PAD 資料區塊可能包含一個 PAD 決策。
1 and 2 1 和 2
0
Y
Y
P7
ISO/IEC 39794-1
A PAD data block may contain PAD score blocks. PAD 資料區塊可包含 PAD 分數區塊。
1 and 2 1 和 2
0
Y
Y
P8
ISO/IEC 39794-1
A PAD data block may contain extended data blocks. 一個 PAD 資料區塊可能包含擴展資料區塊。
1 and 2 1 和 2
0
Y
Y
P9
ISO/IEC 39794-1
A PAD data block may contain a context-of-capture field. 一個 PAD 資料區塊可能包含一個擷取情境欄位。
1 and 2 1 和 2
0
Y
Y
P10
ISO/IEC 39794-1
A PAD data block may contain a level-ofsupervision/surveillance field. PAD 資料區塊可包含監督/監控等級欄位。
1 and 2 1 和 2
0
Y
Y
P11
ISO/IEC 39794-1
A PAD data block may contain a risk level field. PAD 資料區塊可包含風險等級欄位。
1 and 2 1 和 2
0
Y
Y
P12
ISO/IEC 39794-1
A PAD data block may contain a category-of-criteria field. PAD 資料區塊可包含準則類別欄位。
1 and 2 1 和 2
0
Y
Y
P13
ISO/IEC 39794-1
A PAD data block may contain a PAD parameters field. PAD 資料區塊可包含 PAD 參數欄位。
1 and 2 1 和 2
0
Y
Y
P14
ISO/IEC 39794-1
A PAD data block may contain PAD challenges. PAD 資料區塊可能包含 PAD 挑戰。
1 and 2 1 和 2
0
Y
Y
P15
ISO/IEC 39794-1
A PAD data block may contain a PAD capture date/time field. PAD 資料區塊可能包含 PAD 擷取日期/時間欄位。
1 and 2 1 和 2
0
Y
Y
P16
Annex A 附錄 A
A representation block may contain a session identifier. 表示區塊可能包含工作階段識別碼。
1 and 2 1 和 2
0
Y
Y
P17
Annex A 附錄 A
A representation block may contain a derived-from identifier. 表示區塊可能包含衍生自識別碼。
1 and 2 1 和 2
0
Y
Y
P18
Annex A 附錄 A
A representation block may contain a capture device block. 一個表示區塊可能包含一個擷取裝置區塊。
1 and 2 1 和 2
0
Y
Y
P19
Annex A 附錄 A
A capture device block may contain a model identifier block. 一個擷取裝置區塊可能包含一個型號識別碼區塊。
1 and 2 1 和 2
0
Y
Y
P20
Annex A 附錄 A
A capture device block may contain certification identifier blocks. 一個擷取裝置區塊可能包含認證識別碼區塊。
1 and 2 1 和 2
0
Y
Y
P21
Annex A 附錄 A
A capture device block may contain unknown extensions. 一個擷取裝置區塊可能包含未知擴充。
1 and 2 1 和 2
0
Y
Y
P22
Annex A 附錄 A
A representation block may contain an identity metadata block. 一個表示區塊可能包含一個識別中繼資料區塊。
1 and 2 1 和 2
0
Y
Y
P23
Annex A 附錄 A
An identity metadata block may contain a gender field. 身分中繼資料區塊可包含性別欄位。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P1 Annex A A face-image data block may contain unknown extensions. 1 and 2 0 Y Y
P2 Annex A A representation block may contain a capture date/time block. 1 and 2 0 Y Y
P3 AnnexA A representation block may contain quality blocks. 1 and 2 0 Y Y
P4 ISO/IEC 39794-1 A quality block may contain unknown extensions. 1 and 2 0 Y Y
P5 Annex A A representation block may contain a PAD data block. 1 and 2 0 Y Y
P6 ISO/IEC 39794-1 A PAD data block may contain a PAD decision. 1 and 2 0 Y Y
P7 ISO/IEC 39794-1 A PAD data block may contain PAD score blocks. 1 and 2 0 Y Y
P8 ISO/IEC 39794-1 A PAD data block may contain extended data blocks. 1 and 2 0 Y Y
P9 ISO/IEC 39794-1 A PAD data block may contain a context-of-capture field. 1 and 2 0 Y Y
P10 ISO/IEC 39794-1 A PAD data block may contain a level-ofsupervision/surveillance field. 1 and 2 0 Y Y
P11 ISO/IEC 39794-1 A PAD data block may contain a risk level field. 1 and 2 0 Y Y
P12 ISO/IEC 39794-1 A PAD data block may contain a category-of-criteria field. 1 and 2 0 Y Y
P13 ISO/IEC 39794-1 A PAD data block may contain a PAD parameters field. 1 and 2 0 Y Y
P14 ISO/IEC 39794-1 A PAD data block may contain PAD challenges. 1 and 2 0 Y Y
P15 ISO/IEC 39794-1 A PAD data block may contain a PAD capture date/time field. 1 and 2 0 Y Y
P16 Annex A A representation block may contain a session identifier. 1 and 2 0 Y Y
P17 Annex A A representation block may contain a derived-from identifier. 1 and 2 0 Y Y
P18 Annex A A representation block may contain a capture device block. 1 and 2 0 Y Y
P19 Annex A A capture device block may contain a model identifier block. 1 and 2 0 Y Y
P20 Annex A A capture device block may contain certification identifier blocks. 1 and 2 0 Y Y
P21 Annex A A capture device block may contain unknown extensions. 1 and 2 0 Y Y
P22 Annex A A representation block may contain an identity metadata block. 1 and 2 0 Y Y
P23 Annex A An identity metadata block may contain a gender field. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P1 | Annex A | A face-image data block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P2 | Annex A | A representation block may contain a capture date/time block. | 1 and 2 | 0 | Y | Y | | | |
| P3 | AnnexA | A representation block may contain quality blocks. | 1 and 2 | 0 | Y | Y | | | |
| P4 | ISO/IEC 39794-1 | A quality block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P5 | Annex A | A representation block may contain a PAD data block. | 1 and 2 | 0 | Y | Y | | | |
| P6 | ISO/IEC 39794-1 | A PAD data block may contain a PAD decision. | 1 and 2 | 0 | Y | Y | | | |
| P7 | ISO/IEC 39794-1 | A PAD data block may contain PAD score blocks. | 1 and 2 | 0 | Y | Y | | | |
| P8 | ISO/IEC 39794-1 | A PAD data block may contain extended data blocks. | 1 and 2 | 0 | Y | Y | | | |
| P9 | ISO/IEC 39794-1 | A PAD data block may contain a context-of-capture field. | 1 and 2 | 0 | Y | Y | | | |
| P10 | ISO/IEC 39794-1 | A PAD data block may contain a level-ofsupervision/surveillance field. | 1 and 2 | 0 | Y | Y | | | |
| P11 | ISO/IEC 39794-1 | A PAD data block may contain a risk level field. | 1 and 2 | 0 | Y | Y | | | |
| P12 | ISO/IEC 39794-1 | A PAD data block may contain a category-of-criteria field. | 1 and 2 | 0 | Y | Y | | | |
| P13 | ISO/IEC 39794-1 | A PAD data block may contain a PAD parameters field. | 1 and 2 | 0 | Y | Y | | | |
| P14 | ISO/IEC 39794-1 | A PAD data block may contain PAD challenges. | 1 and 2 | 0 | Y | Y | | | |
| P15 | ISO/IEC 39794-1 | A PAD data block may contain a PAD capture date/time field. | 1 and 2 | 0 | Y | Y | | | |
| P16 | Annex A | A representation block may contain a session identifier. | 1 and 2 | 0 | Y | Y | | | |
| P17 | Annex A | A representation block may contain a derived-from identifier. | 1 and 2 | 0 | Y | Y | | | |
| P18 | Annex A | A representation block may contain a capture device block. | 1 and 2 | 0 | Y | Y | | | |
| P19 | Annex A | A capture device block may contain a model identifier block. | 1 and 2 | 0 | Y | Y | | | |
| P20 | Annex A | A capture device block may contain certification identifier blocks. | 1 and 2 | 0 | Y | Y | | | |
| P21 | Annex A | A capture device block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P22 | Annex A | A representation block may contain an identity metadata block. | 1 and 2 | 0 | Y | Y | | | |
| P23 | Annex A | An identity metadata block may contain a gender field. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P24
Annex A 附錄 A
An identity metadata block may contain an eye colour field. 身分識別中繼資料區塊可包含眼睛顏色欄位。
1 and 2 1 和 2
0
Y
Y
P25
Annex A 附錄 A
An identity metadata block may contain a hair colour field. 身分中繼資料區塊可包含髮色欄位。
1 and 2 1 和 2
0
Y
Y
P26
AnnexA 附錄 A
An identity metadata block may contain a subject height field. 身分識別中繼資料區塊可包含主體身高欄位。
1 and 2 1 和 2
0
Y
Y
P27
Annex A 附錄 A
An identity metadata block may contain a properties block. 身分識別中繼資料區塊可包含屬性區塊。
1 and 2 1 和 2
0
Y
Y
P28
Annex A 附錄 A
A properties block may contain a glasses field. 屬性區塊可包含眼鏡欄位。
1 and 2 1 和 2
0
Y
Y
P29
AnnexA 附錄 A
A properties block may contain a moustache field. 屬性區塊可包含鬍鬚欄位。
1 and 2 1 和 2
0
Y
Y
P30
Annex A 附錄 A
A properties block may contain a beard field. 屬性區塊可能包含鬍鬚欄位。
1 and 2 1 和 2
0
Y
Y
P31
Annex A 附錄 A
A properties block may contain a teeth-visible field. 屬性區塊可包含一個牙齒可見欄位。
1 and 2 1 和 2
0
Y
Y
P32
AnnexA 附錄 A
A properties block may contain a pupil-or-iris-not-visible field. 屬性區塊可包含一個瞳孔或虹膜不可見欄位。
1 and 2 1 和 2
0
Y
Y
P33
Annex A 附錄 A
A properties block may contain a mouth-open field. 屬性區塊可能包含一個嘴巴張開欄位。
1 and 2 1 和 2
0
Y
Y
P34
Annex A 附錄 A
A properties block may contain a left-eye-patch field. 屬性區塊可包含左眼罩欄位。
1 and 2 1 和 2
0
Y
Y
P35
Annex A 附錄 A
A properties block may contain a right-eye-patch field. 屬性區塊可包含右眼罩欄位。
1 and 2 1 和 2
0
Y
Y
P36
Annex A 附錄 A
A properties block may contain a dark-glasses field. 屬性區塊可包含深色眼鏡欄位。
1 and 2 1 和 2
0
Y
Y
P37
Annex A 附錄 A
A properties block may contain a biometric-absent field. 屬性區塊可能包含生物特徵不存在欄位。
1 and 2 1 和 2
0
Y
Y
P38
Annex A 附錄 A
A properties block may contain a head-coveringspresent field. 屬性區塊可包含一個頭部覆蓋物存在欄位。
1 and 2 1 和 2
0
Y
Y
P39
Annex A 附錄 A
A properties block may contain unknown extensions. 屬性區塊可包含未知擴充。
1 and 2 1 和 2
0
Y
Y
P40
Annex A 附錄 A
An identity metadata block may contain an expression block. 身分中繼資料區塊可包含表情區塊。
1 and 2 1 和 2
0
Y
Y
P41
Annex A 附錄 A
An expression block may contain a neutral field. 表情區塊可包含中性欄位。
1 and 2 1 和 2
0
Y
Y
P42
Annex A 附錄 A
An expression block may contain a smile field. 表情區塊可包含微笑欄位。
1 and 2 1 和 2
0
Y
Y
P43
Annex A 附錄 A
An expression block may contain a raised-eyebrows field. 表情區塊可包含揚眉欄位。
1 and 2 1 和 2
0
Y
Y
P44
Annex A 附錄 A
An expression block may contain an eyes-looking-away-from-the-camera field. 表情區塊可能包含一個「眼睛未注視攝影機」的欄位。
1 and 2 1 和 2
0
Y
Y
P45
Annex A 附錄 A
An expression block may contain a squinting field. 表達區塊可能包含一個斜視場。
1 and 2 1 和 2
0
Y
Y
P46
Annex A 附錄 A
An expression block may contain a frowning field. 表達區塊可能包含一個皺眉場。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P24 Annex A An identity metadata block may contain an eye colour field. 1 and 2 0 Y Y
P25 Annex A An identity metadata block may contain a hair colour field. 1 and 2 0 Y Y
P26 AnnexA An identity metadata block may contain a subject height field. 1 and 2 0 Y Y
P27 Annex A An identity metadata block may contain a properties block. 1 and 2 0 Y Y
P28 Annex A A properties block may contain a glasses field. 1 and 2 0 Y Y
P29 AnnexA A properties block may contain a moustache field. 1 and 2 0 Y Y
P30 Annex A A properties block may contain a beard field. 1 and 2 0 Y Y
P31 Annex A A properties block may contain a teeth-visible field. 1 and 2 0 Y Y
P32 AnnexA A properties block may contain a pupil-or-iris-not-visible field. 1 and 2 0 Y Y
P33 Annex A A properties block may contain a mouth-open field. 1 and 2 0 Y Y
P34 Annex A A properties block may contain a left-eye-patch field. 1 and 2 0 Y Y
P35 Annex A A properties block may contain a right-eye-patch field. 1 and 2 0 Y Y
P36 Annex A A properties block may contain a dark-glasses field. 1 and 2 0 Y Y
P37 Annex A A properties block may contain a biometric-absent field. 1 and 2 0 Y Y
P38 Annex A A properties block may contain a head-coveringspresent field. 1 and 2 0 Y Y
P39 Annex A A properties block may contain unknown extensions. 1 and 2 0 Y Y
P40 Annex A An identity metadata block may contain an expression block. 1 and 2 0 Y Y
P41 Annex A An expression block may contain a neutral field. 1 and 2 0 Y Y
P42 Annex A An expression block may contain a smile field. 1 and 2 0 Y Y
P43 Annex A An expression block may contain a raised-eyebrows field. 1 and 2 0 Y Y
P44 Annex A An expression block may contain an eyes-looking-away-from-the-camera field. 1 and 2 0 Y Y
P45 Annex A An expression block may contain a squinting field. 1 and 2 0 Y Y
P46 Annex A An expression block may contain a frowning field. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P24 | Annex A | An identity metadata block may contain an eye colour field. | 1 and 2 | 0 | Y | Y | | | |
| P25 | Annex A | An identity metadata block may contain a hair colour field. | 1 and 2 | 0 | Y | Y | | | |
| P26 | AnnexA | An identity metadata block may contain a subject height field. | 1 and 2 | 0 | Y | Y | | | |
| P27 | Annex A | An identity metadata block may contain a properties block. | 1 and 2 | 0 | Y | Y | | | |
| P28 | Annex A | A properties block may contain a glasses field. | 1 and 2 | 0 | Y | Y | | | |
| P29 | AnnexA | A properties block may contain a moustache field. | 1 and 2 | 0 | Y | Y | | | |
| P30 | Annex A | A properties block may contain a beard field. | 1 and 2 | 0 | Y | Y | | | |
| P31 | Annex A | A properties block may contain a teeth-visible field. | 1 and 2 | 0 | Y | Y | | | |
| P32 | AnnexA | A properties block may contain a pupil-or-iris-not-visible field. | 1 and 2 | 0 | Y | Y | | | |
| P33 | Annex A | A properties block may contain a mouth-open field. | 1 and 2 | 0 | Y | Y | | | |
| P34 | Annex A | A properties block may contain a left-eye-patch field. | 1 and 2 | 0 | Y | Y | | | |
| P35 | Annex A | A properties block may contain a right-eye-patch field. | 1 and 2 | 0 | Y | Y | | | |
| P36 | Annex A | A properties block may contain a dark-glasses field. | 1 and 2 | 0 | Y | Y | | | |
| P37 | Annex A | A properties block may contain a biometric-absent field. | 1 and 2 | 0 | Y | Y | | | |
| P38 | Annex A | A properties block may contain a head-coveringspresent field. | 1 and 2 | 0 | Y | Y | | | |
| P39 | Annex A | A properties block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P40 | Annex A | An identity metadata block may contain an expression block. | 1 and 2 | 0 | Y | Y | | | |
| P41 | Annex A | An expression block may contain a neutral field. | 1 and 2 | 0 | Y | Y | | | |
| P42 | Annex A | An expression block may contain a smile field. | 1 and 2 | 0 | Y | Y | | | |
| P43 | Annex A | An expression block may contain a raised-eyebrows field. | 1 and 2 | 0 | Y | Y | | | |
| P44 | Annex A | An expression block may contain an eyes-looking-away-from-the-camera field. | 1 and 2 | 0 | Y | Y | | | |
| P45 | Annex A | An expression block may contain a squinting field. | 1 and 2 | 0 | Y | Y | | | |
| P46 | Annex A | An expression block may contain a frowning field. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P47
Annex A 附錄 A
An expression block may contain unknown extensions. 表達式區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P48
Annex A 附錄 A
An identity metadata block may contain a pose angle block. 身分中繼資料區塊可能包含姿態角度區塊。
1 and 2 1 和 2
0
Y
Y
P49
Annex A 附錄 A
A pose angle block may contain a yaw angle block. 一個姿態角度區塊可能包含一個偏航角度區塊。
1 and 2 1 和 2
0
Y
Y
P50
Annex A 附錄 A
A yaw angle block may contain an angle value field. 偏航角區塊可能包含角度值欄位。
1 and 2 1 和 2
0
Y
Y
P51
Annex A 附錄 A
A yaw angle block may contain an angle uncertainty field. 偏航角區塊可能包含角度不確定性欄位。
1 and 2 1 和 2
0
Y
Y
P52
Annex A 附錄 A
A yaw angle block may contain unknown extensions. 偏航角區塊可能包含未知擴充。
1 and 2 1 和 2
0
Y
Y
P53
Annex A 附錄 A
A pose angle block may contain a pitch angle block. 姿態角度區塊可包含俯仰角度區塊。
1 and 2 1 和 2
0
Y
Y
P54
Annex A 附錄 A
A pitch angle block may contain an angle value field. 俯仰角度區塊可包含角度值欄位。
1 and 2 1 和 2
0
Y
Y
P55
Annex A 附錄 A
A pitch angle block may contain an angle uncertainty field. 俯仰角度區塊可包含角度不確定性欄位。
1 and 2 1 和 2
0
Y
Y
P56
Annex A 附錄 A
A pitch angle block may contain unknown extensions. 俯仰角區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P57
Annex A 附錄 A
A pose angle block may contain a roll angle block. 姿態角區塊可能包含滾轉角區塊。
1 and 2 1 和 2
0
Y
Y
P58
Annex A 附錄 A
A roll angle block may contain an angle value field. 滾轉角區塊可能包含角度值欄位。
1 and 2 1 和 2
0
Y
Y
P59
Annex A 附錄 A
A roll angle block may contain an angle uncertainty field. 滾轉角區塊可能包含角度不確定性欄位。
1 and 2 1 和 2
0
Y
Y
P60
Annex A 附錄 A
A roll angle block may contain unknown extensions. 翻滾角區塊可能包含不明的擴充功能。
1 and 2 1 和 2
0
Y
Y
P61
Annex A 附錄 A
An identity metadata block may contain unknown extensions. 身分識別中繼資料區塊可能包含不明的擴充功能。
1 and 2 1 和 2
0
Y
Y
P62
Annex A 附錄 A
A representation block may contain landmarks blocks. 一個表示區塊可能包含地標區塊。
1 and 2 1 和 2
0
Y
Y
P63
Annex A 附錄 A
A landmark block may contain a landmark kind value. 地標區塊可能包含地標種類值。
1 and 2 1 和 2
0
Y
Y
P64
AnnexA 附錄 A
A landmark block may contain a landmark coordinates block. 地標區塊可包含地標座標區塊。
1 and 2 1 和 2
0
Y
Y
P65
Annex A 附錄 A
A landmark co-ordinates block may contain a 2D Cartesian coordinates block. 地標座標區塊可包含二維笛卡兒座標區塊。
1 and 2 1 和 2
0
Y
Y
P66
Annex A 附錄 A
A landmark co-ordinates block may contain a texture image coordinates block. 地標座標區塊可包含紋理影像座標區塊。
1 and 2 1 和 2
0
Y
Y
P67
Annex A 附錄 A
A landmark coordinates block may contain a 3D Cartesian coordinates block. 地標座標區塊可包含三維笛卡兒座標區塊。
1 and 2 1 和 2
0
Y
Y
P68
Annex A 附錄 A
A landmark coordinates block may contain unknown extensions. 地標座標區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P69
Annex A 附錄 A
A landmark block may contain unknown extensions. 地標區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P47 Annex A An expression block may contain unknown extensions. 1 and 2 0 Y Y
P48 Annex A An identity metadata block may contain a pose angle block. 1 and 2 0 Y Y
P49 Annex A A pose angle block may contain a yaw angle block. 1 and 2 0 Y Y
P50 Annex A A yaw angle block may contain an angle value field. 1 and 2 0 Y Y
P51 Annex A A yaw angle block may contain an angle uncertainty field. 1 and 2 0 Y Y
P52 Annex A A yaw angle block may contain unknown extensions. 1 and 2 0 Y Y
P53 Annex A A pose angle block may contain a pitch angle block. 1 and 2 0 Y Y
P54 Annex A A pitch angle block may contain an angle value field. 1 and 2 0 Y Y
P55 Annex A A pitch angle block may contain an angle uncertainty field. 1 and 2 0 Y Y
P56 Annex A A pitch angle block may contain unknown extensions. 1 and 2 0 Y Y
P57 Annex A A pose angle block may contain a roll angle block. 1 and 2 0 Y Y
P58 Annex A A roll angle block may contain an angle value field. 1 and 2 0 Y Y
P59 Annex A A roll angle block may contain an angle uncertainty field. 1 and 2 0 Y Y
P60 Annex A A roll angle block may contain unknown extensions. 1 and 2 0 Y Y
P61 Annex A An identity metadata block may contain unknown extensions. 1 and 2 0 Y Y
P62 Annex A A representation block may contain landmarks blocks. 1 and 2 0 Y Y
P63 Annex A A landmark block may contain a landmark kind value. 1 and 2 0 Y Y
P64 AnnexA A landmark block may contain a landmark coordinates block. 1 and 2 0 Y Y
P65 Annex A A landmark co-ordinates block may contain a 2D Cartesian coordinates block. 1 and 2 0 Y Y
P66 Annex A A landmark co-ordinates block may contain a texture image coordinates block. 1 and 2 0 Y Y
P67 Annex A A landmark coordinates block may contain a 3D Cartesian coordinates block. 1 and 2 0 Y Y
P68 Annex A A landmark coordinates block may contain unknown extensions. 1 and 2 0 Y Y
P69 Annex A A landmark block may contain unknown extensions. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P47 | Annex A | An expression block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P48 | Annex A | An identity metadata block may contain a pose angle block. | 1 and 2 | 0 | Y | Y | | | |
| P49 | Annex A | A pose angle block may contain a yaw angle block. | 1 and 2 | 0 | Y | Y | | | |
| P50 | Annex A | A yaw angle block may contain an angle value field. | 1 and 2 | 0 | Y | Y | | | |
| P51 | Annex A | A yaw angle block may contain an angle uncertainty field. | 1 and 2 | 0 | Y | Y | | | |
| P52 | Annex A | A yaw angle block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P53 | Annex A | A pose angle block may contain a pitch angle block. | 1 and 2 | 0 | Y | Y | | | |
| P54 | Annex A | A pitch angle block may contain an angle value field. | 1 and 2 | 0 | Y | Y | | | |
| P55 | Annex A | A pitch angle block may contain an angle uncertainty field. | 1 and 2 | 0 | Y | Y | | | |
| P56 | Annex A | A pitch angle block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P57 | Annex A | A pose angle block may contain a roll angle block. | 1 and 2 | 0 | Y | Y | | | |
| P58 | Annex A | A roll angle block may contain an angle value field. | 1 and 2 | 0 | Y | Y | | | |
| P59 | Annex A | A roll angle block may contain an angle uncertainty field. | 1 and 2 | 0 | Y | Y | | | |
| P60 | Annex A | A roll angle block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P61 | Annex A | An identity metadata block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P62 | Annex A | A representation block may contain landmarks blocks. | 1 and 2 | 0 | Y | Y | | | |
| P63 | Annex A | A landmark block may contain a landmark kind value. | 1 and 2 | 0 | Y | Y | | | |
| P64 | AnnexA | A landmark block may contain a landmark coordinates block. | 1 and 2 | 0 | Y | Y | | | |
| P65 | Annex A | A landmark co-ordinates block may contain a 2D Cartesian coordinates block. | 1 and 2 | 0 | Y | Y | | | |
| P66 | Annex A | A landmark co-ordinates block may contain a texture image coordinates block. | 1 and 2 | 0 | Y | Y | | | |
| P67 | Annex A | A landmark coordinates block may contain a 3D Cartesian coordinates block. | 1 and 2 | 0 | Y | Y | | | |
| P68 | Annex A | A landmark coordinates block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P69 | Annex A | A landmark block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P70
Annex A 附錄 A
A representation block may contain unknown extensions. 一個表示區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P71
Annex A 附錄 A
An image representation may contain a 2D image representation block. 影像表示可包含二維影像表示區塊。
1 and 2 1 和 2
0
Y
Y
P72
Annex A 附錄 A
A 2D image representation block may contain a 2D capture device block. 二維影像表示區塊可能包含二維擷取裝置區塊。
1 and 2 1 和 2
0
Y
Y
P73
Annex A 附錄 A
A 2D capture device block may contain a 2D capture device spectral block. 二維擷取裝置區塊可包含二維擷取裝置光譜區塊。
1 and 2 1 和 2
0
Y
Y
P74
Annex A 附錄 A
A 2D capture device spectral block may contain a white-light field. 二維擷取裝置光譜區塊可包含白光場。
1 and 2 1 和 2
0
Y
Y
P75
Annex A 附錄 A
A2D capture device spectral block may contain a near-infrared field. 二維擷取裝置光譜區塊可包含近紅外線場。
1 and 2 1 和 2
0
Y
Y
P76
Annex A 附錄 A
A 2D capture device spectral block may contain a thermal field. 二維擷取裝置光譜區塊可包含熱場。
1 and 2 1 和 2
0
Y
Y
P77
Annex A 附錄 A
A 2D capture device spectral block may contain unknown extensions. 2D 擷取裝置光譜區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P78
Annex A 附錄 A
A 2D capture device block may contain a 2D capture device technology identifier field. 二維擷取裝置區塊可能包含二維擷取裝置技術識別碼欄位。
1 and 2 1 和 2
0
Y
Y
P79
Annex A 附錄 A
A 2D capture device block may contain unknown extensions. 2D 擷取裝置區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P80
Annex A 附錄 A
A 2D image information block may contain a 2D face image kind field. 二維影像資訊區塊可能包含二維人臉影像類型欄位。
1 and 2 1 和 2
0
Y
Y
P81
Annex A 附錄 A
A 2D image information block may contain a postacquisition processing block. 二維影像資訊區塊可能包含後處理區塊。
1 and 2 1 和 2
0
Y
Y
P82
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a rotated field. 二維影像資訊區塊中的後處理區塊可能包含旋轉場。
1 and 2 1 和 2
0
Y
Y
P83
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a cropped field. 二維影像資訊區塊中的後擷取處理區塊可能包含裁切欄位。
1 and 2 1 和 2
0
Y
Y
P84
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a down-sampled field. 二維影像資訊區塊中的後擷取處理區塊可能包含降取樣欄位。
1 and 2 1 和 2
0
Y
Y
P85
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a white-balance-adjusted field. 二維影像資訊區塊中的後擷取處理區塊可包含白平衡調整欄位。
1 and 2 1 和 2
0
Y
Y
P86
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a multiplycompressed field. 二維影像資訊區塊中的後擷取處理區塊可包含多重壓縮欄位。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P70 Annex A A representation block may contain unknown extensions. 1 and 2 0 Y Y
P71 Annex A An image representation may contain a 2D image representation block. 1 and 2 0 Y Y
P72 Annex A A 2D image representation block may contain a 2D capture device block. 1 and 2 0 Y Y
P73 Annex A A 2D capture device block may contain a 2D capture device spectral block. 1 and 2 0 Y Y
P74 Annex A A 2D capture device spectral block may contain a white-light field. 1 and 2 0 Y Y
P75 Annex A A2D capture device spectral block may contain a near-infrared field. 1 and 2 0 Y Y
P76 Annex A A 2D capture device spectral block may contain a thermal field. 1 and 2 0 Y Y
P77 Annex A A 2D capture device spectral block may contain unknown extensions. 1 and 2 0 Y Y
P78 Annex A A 2D capture device block may contain a 2D capture device technology identifier field. 1 and 2 0 Y Y
P79 Annex A A 2D capture device block may contain unknown extensions. 1 and 2 0 Y Y
P80 Annex A A 2D image information block may contain a 2D face image kind field. 1 and 2 0 Y Y
P81 Annex A A 2D image information block may contain a postacquisition processing block. 1 and 2 0 Y Y
P82 Annex A A post-acquisition processing block within a 2D image information block may contain a rotated field. 1 and 2 0 Y Y
P83 Annex A A post-acquisition processing block within a 2D image information block may contain a cropped field. 1 and 2 0 Y Y
P84 Annex A A post-acquisition processing block within a 2D image information block may contain a down-sampled field. 1 and 2 0 Y Y
P85 Annex A A post-acquisition processing block within a 2D image information block may contain a white-balance-adjusted field. 1 and 2 0 Y Y
P86 Annex A A post-acquisition processing block within a 2D image information block may contain a multiplycompressed field. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P70 | Annex A | A representation block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P71 | Annex A | An image representation may contain a 2D image representation block. | 1 and 2 | 0 | Y | Y | | | |
| P72 | Annex A | A 2D image representation block may contain a 2D capture device block. | 1 and 2 | 0 | Y | Y | | | |
| P73 | Annex A | A 2D capture device block may contain a 2D capture device spectral block. | 1 and 2 | 0 | Y | Y | | | |
| P74 | Annex A | A 2D capture device spectral block may contain a white-light field. | 1 and 2 | 0 | Y | Y | | | |
| P75 | Annex A | A2D capture device spectral block may contain a near-infrared field. | 1 and 2 | 0 | Y | Y | | | |
| P76 | Annex A | A 2D capture device spectral block may contain a thermal field. | 1 and 2 | 0 | Y | Y | | | |
| P77 | Annex A | A 2D capture device spectral block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P78 | Annex A | A 2D capture device block may contain a 2D capture device technology identifier field. | 1 and 2 | 0 | Y | Y | | | |
| P79 | Annex A | A 2D capture device block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P80 | Annex A | A 2D image information block may contain a 2D face image kind field. | 1 and 2 | 0 | Y | Y | | | |
| P81 | Annex A | A 2D image information block may contain a postacquisition processing block. | 1 and 2 | 0 | Y | Y | | | |
| P82 | Annex A | A post-acquisition processing block within a 2D image information block may contain a rotated field. | 1 and 2 | 0 | Y | Y | | | |
| P83 | Annex A | A post-acquisition processing block within a 2D image information block may contain a cropped field. | 1 and 2 | 0 | Y | Y | | | |
| P84 | Annex A | A post-acquisition processing block within a 2D image information block may contain a down-sampled field. | 1 and 2 | 0 | Y | Y | | | |
| P85 | Annex A | A post-acquisition processing block within a 2D image information block may contain a white-balance-adjusted field. | 1 and 2 | 0 | Y | Y | | | |
| P86 | Annex A | A post-acquisition processing block within a 2D image information block may contain a multiplycompressed field. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P87
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain an interpolated field. 二維影像資訊區塊中的後處理區塊可能包含內插欄位。
1 and 2 1 和 2
0
Y
Y
P88
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a contrast-stretched field. 二維影像資訊區塊中的後處理區塊可能包含對比度拉伸欄位。
1 and 2 1 和 2
0
Y
Y
P89
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a pose-corrected field. 二維影像資訊區塊中的後處理區塊可能包含姿勢校正欄位。
1 and 2 1 和 2
0
Y
Y
P90
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a multi-view image field. 二維影像資訊區塊中的後處理區塊可包含多視角影像欄位。
1 and 2 1 和 2
0
Y
Y
P91
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain an age-progressed field. 二維影像資訊區塊中的後處理區塊可包含年齡增長欄位。
1 and 2 1 和 2
0
Y
Y
P92
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a super-resolution processed field. 二維影像資訊區塊中的後處理區塊可包含超解析度處理欄位。
1 and 2 1 和 2
0
Y
Y
P93
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain a normalised field. 二維影像資訊區塊中的後處理區塊可包含正規化欄位。
1 and 2 1 和 2
0
Y
Y
P94
Annex A 附錄 A
A post-acquisition processing block within a 2D image information block may contain unknown extensions. 二維影像資訊區塊中的後擷取處理區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P95
Annex A 附錄 A
A 2D image information block may contain a lossytransformation attempts field. 二維影像資訊區塊可能包含有損轉換嘗試欄位。
1 and 2 1 和 2
0
Y
Y
P96
Annex A 附錄 A
A 2D image information block may contain a camera-to-subject distance field. 二維影像資訊區塊可能包含相機到主體的距離欄位。
1 and 2 1 和 2
0
Y
Y
P97
Annex A 附錄 A
A 2D image information block may contain sensor diagonal field. 二維影像資訊區塊可能包含感測器對角線場。
1 and 2 1 和 2
0
Y
Y
P98
Annex A 附錄 A
A 2D image information block may contain a lens focal length field. 二維影像資訊區塊可包含鏡頭焦距欄位。
1 and 2 1 和 2
0
Y
Y
P99
Annex A 附錄 A
A 2D image information block may contain an image size block. 二維影像資訊區塊可包含影像尺寸區塊。
1 and 2 1 和 2
0
Y
Y
P100
Annex A 附錄 A
If the 2D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. 如果二維影像資料格式未知、為其他格式或為較新版本的擴充功能代碼,則應包含影像尺寸區塊(寬度和高度)。
1 and 2 1 和 2
0
Y
Y
P101
Annex A 附錄 A
An image size block may contain a width field. 影像尺寸區塊可包含寬度欄位。
1 and 2 1 和 2
0
Y
Y
P102
Annex A 附錄 A
An image size block may contain a height field. 影像尺寸區塊可能包含高度欄位。
1 and 2 1 和 2
0
Y
Y
P103
Annex A 附錄 A
A 2D image information block may contain an image face measurements block. 二維影像資訊區塊可能包含影像臉部測量區塊。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P87 Annex A A post-acquisition processing block within a 2D image information block may contain an interpolated field. 1 and 2 0 Y Y
P88 Annex A A post-acquisition processing block within a 2D image information block may contain a contrast-stretched field. 1 and 2 0 Y Y
P89 Annex A A post-acquisition processing block within a 2D image information block may contain a pose-corrected field. 1 and 2 0 Y Y
P90 Annex A A post-acquisition processing block within a 2D image information block may contain a multi-view image field. 1 and 2 0 Y Y
P91 Annex A A post-acquisition processing block within a 2D image information block may contain an age-progressed field. 1 and 2 0 Y Y
P92 Annex A A post-acquisition processing block within a 2D image information block may contain a super-resolution processed field. 1 and 2 0 Y Y
P93 Annex A A post-acquisition processing block within a 2D image information block may contain a normalised field. 1 and 2 0 Y Y
P94 Annex A A post-acquisition processing block within a 2D image information block may contain unknown extensions. 1 and 2 0 Y Y
P95 Annex A A 2D image information block may contain a lossytransformation attempts field. 1 and 2 0 Y Y
P96 Annex A A 2D image information block may contain a camera-to-subject distance field. 1 and 2 0 Y Y
P97 Annex A A 2D image information block may contain sensor diagonal field. 1 and 2 0 Y Y
P98 Annex A A 2D image information block may contain a lens focal length field. 1 and 2 0 Y Y
P99 Annex A A 2D image information block may contain an image size block. 1 and 2 0 Y Y
P100 Annex A If the 2D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. 1 and 2 0 Y Y
P101 Annex A An image size block may contain a width field. 1 and 2 0 Y Y
P102 Annex A An image size block may contain a height field. 1 and 2 0 Y Y
P103 Annex A A 2D image information block may contain an image face measurements block. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P87 | Annex A | A post-acquisition processing block within a 2D image information block may contain an interpolated field. | 1 and 2 | 0 | Y | Y | | | |
| P88 | Annex A | A post-acquisition processing block within a 2D image information block may contain a contrast-stretched field. | 1 and 2 | 0 | Y | Y | | | |
| P89 | Annex A | A post-acquisition processing block within a 2D image information block may contain a pose-corrected field. | 1 and 2 | 0 | Y | Y | | | |
| P90 | Annex A | A post-acquisition processing block within a 2D image information block may contain a multi-view image field. | 1 and 2 | 0 | Y | Y | | | |
| P91 | Annex A | A post-acquisition processing block within a 2D image information block may contain an age-progressed field. | 1 and 2 | 0 | Y | Y | | | |
| P92 | Annex A | A post-acquisition processing block within a 2D image information block may contain a super-resolution processed field. | 1 and 2 | 0 | Y | Y | | | |
| P93 | Annex A | A post-acquisition processing block within a 2D image information block may contain a normalised field. | 1 and 2 | 0 | Y | Y | | | |
| P94 | Annex A | A post-acquisition processing block within a 2D image information block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P95 | Annex A | A 2D image information block may contain a lossytransformation attempts field. | 1 and 2 | 0 | Y | Y | | | |
| P96 | Annex A | A 2D image information block may contain a camera-to-subject distance field. | 1 and 2 | 0 | Y | Y | | | |
| P97 | Annex A | A 2D image information block may contain sensor diagonal field. | 1 and 2 | 0 | Y | Y | | | |
| P98 | Annex A | A 2D image information block may contain a lens focal length field. | 1 and 2 | 0 | Y | Y | | | |
| P99 | Annex A | A 2D image information block may contain an image size block. | 1 and 2 | 0 | Y | Y | | | |
| P100 | Annex A | If the 2D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. | 1 and 2 | 0 | Y | Y | | | |
| P101 | Annex A | An image size block may contain a width field. | 1 and 2 | 0 | Y | Y | | | |
| P102 | Annex A | An image size block may contain a height field. | 1 and 2 | 0 | Y | Y | | | |
| P103 | Annex A | A 2D image information block may contain an image face measurements block. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P104
Annex A 附錄 A
An image face measurements block may contain an image headwidth field. 影像臉部測量區塊可包含影像頭部寬度欄位。
1 and 2 1 和 2
0
Y
Y
P105
Annex A 附錄 A
An image face measurements block may contain an image in-ter-eye distance field. 影像臉部測量區塊可包含影像雙眼間距欄位。
1 and 2 1 和 2
0
Y
Y
P106
Annex A 附錄 A
An image face measurements block may contain an image eye-to-mouth distance field. 影像臉部測量區塊可包含影像眼部至嘴部距離欄位。
1 and 2 1 和 2
0
Y
Y
P107
Annex A 附錄 A
An image face measurements block may contain an image headlength field. 影像臉部測量區塊可包含影像頭部長度欄位。
1 and 2 1 和 2
0
Y
Y
P108
Annex A 附錄 A
An image face measurements block may contain unknown extensions. 影像臉部測量區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P109
Annex A 附錄 A
A 2D image information block may contain an image colour space field. 二維影像資訊區塊可能包含影像色彩空間欄位。
1 and 2 1 和 2
0
Y
Y
P110
Annex A 附錄 A
A 2D image information block may contain a reference colour mapping block. 二維影像資訊區塊可能包含參考色彩對應區塊。
1 and 2 1 和 2
0
Y
Y
P111
Annex A 附錄 A
A reference colour mapping block may contain a reference colour scheme field. 參考色彩對應區塊可包含參考色彩配置欄位。
1 and 2 1 和 2
0
Y
Y
P112
Annex A 附錄 A
A reference colour mapping block may contain reference colour definition-and-value blocks. 參考色彩對應區塊可包含參考色彩定義與數值區塊。
1 and 2 1 和 2
0
Y
Y
P113
Annex A 附錄 A
A reference colour definition-and-value block may contain a reference colour definition field. 參考色彩定義與數值區塊可能包含參考色彩定義欄位。
1 and 2 1 和 2
0
Y
Y
P114
Annex A 附錄 A
A reference colour definition-and-value block may contain a reference colour value field. 參考色彩定義與數值區塊可包含參考色彩數值欄位。
1 and 2 1 和 2
0
Y
Y
P115
Annex A 附錄 A
A reference colour definition-and-value block may contain unknown extensions. 參考色彩定義與數值區塊可包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P116
Annex A 附錄 A
A reference colour mapping block may contain unknown extensions. 參考色彩對應區塊可包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P117
Annex A 附錄 A
A 2D image information block may contain unknown extensions. 2D 影像資訊區塊可包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P118
Annex A 附錄 A
A 2D image representation block may contain unknown extensions. 二維影像表示區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P119
Annex A 附錄 A
An image representation may contain a 3D shape representation block. 影像表示法可能包含 3D 形狀表示區塊。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P104 Annex A An image face measurements block may contain an image headwidth field. 1 and 2 0 Y Y
P105 Annex A An image face measurements block may contain an image in-ter-eye distance field. 1 and 2 0 Y Y
P106 Annex A An image face measurements block may contain an image eye-to-mouth distance field. 1 and 2 0 Y Y
P107 Annex A An image face measurements block may contain an image headlength field. 1 and 2 0 Y Y
P108 Annex A An image face measurements block may contain unknown extensions. 1 and 2 0 Y Y
P109 Annex A A 2D image information block may contain an image colour space field. 1 and 2 0 Y Y
P110 Annex A A 2D image information block may contain a reference colour mapping block. 1 and 2 0 Y Y
P111 Annex A A reference colour mapping block may contain a reference colour scheme field. 1 and 2 0 Y Y
P112 Annex A A reference colour mapping block may contain reference colour definition-and-value blocks. 1 and 2 0 Y Y
P113 Annex A A reference colour definition-and-value block may contain a reference colour definition field. 1 and 2 0 Y Y
P114 Annex A A reference colour definition-and-value block may contain a reference colour value field. 1 and 2 0 Y Y
P115 Annex A A reference colour definition-and-value block may contain unknown extensions. 1 and 2 0 Y Y
P116 Annex A A reference colour mapping block may contain unknown extensions. 1 and 2 0 Y Y
P117 Annex A A 2D image information block may contain unknown extensions. 1 and 2 0 Y Y
P118 Annex A A 2D image representation block may contain unknown extensions. 1 and 2 0 Y Y
P119 Annex A An image representation may contain a 3D shape representation block. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P104 | Annex A | An image face measurements block may contain an image headwidth field. | 1 and 2 | 0 | Y | Y | | | |
| P105 | Annex A | An image face measurements block may contain an image in-ter-eye distance field. | 1 and 2 | 0 | Y | Y | | | |
| P106 | Annex A | An image face measurements block may contain an image eye-to-mouth distance field. | 1 and 2 | 0 | Y | Y | | | |
| P107 | Annex A | An image face measurements block may contain an image headlength field. | 1 and 2 | 0 | Y | Y | | | |
| P108 | Annex A | An image face measurements block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P109 | Annex A | A 2D image information block may contain an image colour space field. | 1 and 2 | 0 | Y | Y | | | |
| P110 | Annex A | A 2D image information block may contain a reference colour mapping block. | 1 and 2 | 0 | Y | Y | | | |
| P111 | Annex A | A reference colour mapping block may contain a reference colour scheme field. | 1 and 2 | 0 | Y | Y | | | |
| P112 | Annex A | A reference colour mapping block may contain reference colour definition-and-value blocks. | 1 and 2 | 0 | Y | Y | | | |
| P113 | Annex A | A reference colour definition-and-value block may contain a reference colour definition field. | 1 and 2 | 0 | Y | Y | | | |
| P114 | Annex A | A reference colour definition-and-value block may contain a reference colour value field. | 1 and 2 | 0 | Y | Y | | | |
| P115 | Annex A | A reference colour definition-and-value block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P116 | Annex A | A reference colour mapping block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P117 | Annex A | A 2D image information block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P118 | Annex A | A 2D image representation block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P119 | Annex A | An image representation may contain a 3D shape representation block. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P120
Annex A 附錄 A
A 3D shape representation block may contain a 3D capture device block. 3D 形狀表示區塊可包含 3D 擷取裝置區塊。
1 and 2 1 和 2
0
Y
Y
P121
Annex A 附錄 A
A 3D capture device technology block may contain a 3D modus field. 3D 擷取裝置技術區塊可能包含 3D 模式欄位。
1 and 2 1 和 2
0
Y
Y
P122
Annex A 附錄 A
A 3D capture device technology block may contain a 3D capture device technology kind field. 3D 擷取裝置技術區塊可能包含 3D 擷取裝置技術種類欄位。
1 and 2 1 和 2
0
Y
Y
P123
Annex A 附錄 A
A 3D capture device technology block may contain unknown extensions. 3D 擷取裝置技術區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P124
Annex A 附錄 A
A 3D image information block may contain an image colour space field. 3D 影像資訊區塊可能包含影像色彩空間欄位。
1 and 2 1 和 2
0
Y
Y
P125
Annex A 附錄 A
A 3D image information block may contain a 3D face image kind field. 一個 3D 影像資訊區塊可包含一個 3D 臉部影像種類欄位。
1 and 2 1 和 2
0
Y
Y
P126
Annex A 附錄 A
A 3D image information block may contain an image size block. 一個 3D 影像資訊區塊可包含一個影像尺寸區塊。
1 and 2 1 和 2
0
Y
Y
P127
Annex A 附錄 A
If the 3D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. 如果 3D 影像資料格式為未知、其他或較新版本擴充功能代碼,則應包含一個影像尺寸區塊(寬度和高度)。
1 and 2 1 和 2
0
Y
Y
P128
Annex A 附錄 A
A 3D image information block may contain a 3D physical face measurements block. 一個 3D 影像資訊區塊可包含一個 3D 實體臉部測量區塊。
1 and 2 1 和 2
0
Y
Y
P129
Annex A 附錄 A
A physical face measurements block may contain a 3D physical head-width field. 實體臉部測量區塊可包含 3D 實體頭部寬度欄位。
1 and 2 1 和 2
0
Y
Y
P130
Annex A 附錄 A
A physical face measurements block may contain a 3D physical inter-eye distance field. 實體臉部測量區塊可包含 3D 實體眼距欄位。
1 and 2 1 和 2
0
Y
Y
P131
Annex A 附錄 A
A physical face measurements block may contain a 3D physical eye-to-mouth distance field. 實體臉部測量區塊可包含 3D 實體眼到嘴距離欄位。
1 and 2 1 和 2
0
Y
Y
P132
Annex A 附錄 A
A physical face measurements block may contain a 3D physical head-length field. 實體臉部測量區塊可包含 3D 實體頭部長度欄位。
1 and 2 1 和 2
0
Y
Y
P133
Annex A 附錄 A
A physical face measurements block may contain unknown extensions. 實體臉部測量區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P134
Annex A 附錄 A
A 3D image information block may contain a postacquisition processing block. 3D 影像資訊區塊可能包含後處理區塊。
1 and 2 1 和 2
0
Y
Y
P135
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a rotated field. 3D 影像資訊區塊內的後處理區塊可能包含旋轉場。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P120 Annex A A 3D shape representation block may contain a 3D capture device block. 1 and 2 0 Y Y
P121 Annex A A 3D capture device technology block may contain a 3D modus field. 1 and 2 0 Y Y
P122 Annex A A 3D capture device technology block may contain a 3D capture device technology kind field. 1 and 2 0 Y Y
P123 Annex A A 3D capture device technology block may contain unknown extensions. 1 and 2 0 Y Y
P124 Annex A A 3D image information block may contain an image colour space field. 1 and 2 0 Y Y
P125 Annex A A 3D image information block may contain a 3D face image kind field. 1 and 2 0 Y Y
P126 Annex A A 3D image information block may contain an image size block. 1 and 2 0 Y Y
P127 Annex A If the 3D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. 1 and 2 0 Y Y
P128 Annex A A 3D image information block may contain a 3D physical face measurements block. 1 and 2 0 Y Y
P129 Annex A A physical face measurements block may contain a 3D physical head-width field. 1 and 2 0 Y Y
P130 Annex A A physical face measurements block may contain a 3D physical inter-eye distance field. 1 and 2 0 Y Y
P131 Annex A A physical face measurements block may contain a 3D physical eye-to-mouth distance field. 1 and 2 0 Y Y
P132 Annex A A physical face measurements block may contain a 3D physical head-length field. 1 and 2 0 Y Y
P133 Annex A A physical face measurements block may contain unknown extensions. 1 and 2 0 Y Y
P134 Annex A A 3D image information block may contain a postacquisition processing block. 1 and 2 0 Y Y
P135 Annex A A post-acquisition processing block within a 3D image information block may contain a rotated field. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P120 | Annex A | A 3D shape representation block may contain a 3D capture device block. | 1 and 2 | 0 | Y | Y | | | |
| P121 | Annex A | A 3D capture device technology block may contain a 3D modus field. | 1 and 2 | 0 | Y | Y | | | |
| P122 | Annex A | A 3D capture device technology block may contain a 3D capture device technology kind field. | 1 and 2 | 0 | Y | Y | | | |
| P123 | Annex A | A 3D capture device technology block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P124 | Annex A | A 3D image information block may contain an image colour space field. | 1 and 2 | 0 | Y | Y | | | |
| P125 | Annex A | A 3D image information block may contain a 3D face image kind field. | 1 and 2 | 0 | Y | Y | | | |
| P126 | Annex A | A 3D image information block may contain an image size block. | 1 and 2 | 0 | Y | Y | | | |
| P127 | Annex A | If the 3D image data format is unknown or other or a later version extension code, then an image size block (width and height) shall be included. | 1 and 2 | 0 | Y | Y | | | |
| P128 | Annex A | A 3D image information block may contain a 3D physical face measurements block. | 1 and 2 | 0 | Y | Y | | | |
| P129 | Annex A | A physical face measurements block may contain a 3D physical head-width field. | 1 and 2 | 0 | Y | Y | | | |
| P130 | Annex A | A physical face measurements block may contain a 3D physical inter-eye distance field. | 1 and 2 | 0 | Y | Y | | | |
| P131 | Annex A | A physical face measurements block may contain a 3D physical eye-to-mouth distance field. | 1 and 2 | 0 | Y | Y | | | |
| P132 | Annex A | A physical face measurements block may contain a 3D physical head-length field. | 1 and 2 | 0 | Y | Y | | | |
| P133 | Annex A | A physical face measurements block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P134 | Annex A | A 3D image information block may contain a postacquisition processing block. | 1 and 2 | 0 | Y | Y | | | |
| P135 | Annex A | A post-acquisition processing block within a 3D image information block may contain a rotated field. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P136
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a cropped field. 3D 影像資訊區塊中的後擷取處理區塊可包含裁切欄位。
1 and 2 1 和 2
0
Y
Y
P137
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a down-sampled field. 3D 影像資訊區塊中的後擷取處理區塊可包含降取樣欄位。
1 and 2 1 和 2
0
Y
Y
P138
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a white-balance-adjusted field. 3D 影像資訊區塊中的後擷取處理區塊可包含白平衡調整欄位。
1 and 2 1 和 2
0
Y
Y
P139
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a multiply-compressed field. 3D 影像資訊區塊中的後擷取處理區塊可包含多重壓縮欄位。
1 and 2 1 和 2
0
Y
Y
P140
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain an interpolated field. 3D 影像資訊區塊中的後擷取處理區塊可包含內插欄位。
1 and 2 1 和 2
0
Y
Y
P141
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a contrast-stretched field. 3D 影像資訊區塊中的後擷取處理區塊可包含對比度拉伸欄位。
1 and 2 1 和 2
0
Y
Y
P142
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a pose-corrected field. 3D 影像資訊區塊中的後擷取處理區塊可包含姿勢校正欄位。
1 and 2 1 和 2
0
Y
Y
P143
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a multi-view image field. 3D 影像資訊區塊中的後擷取處理區塊可包含多視角影像欄位。
1 and 2 1 和 2
0
Y
Y
P144
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain an age-progressed field. 3D 影像資訊區塊中的後擷取處理區塊可包含年齡進展欄位。
1 and 2 1 和 2
0
Y
Y
P145
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a super-resolution processed field. 三維影像資訊區塊中的後擷取處理區塊可包含超解析度處理欄位。
1 and 2 1 和 2
0
Y
Y
P146
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain a normalised field. 三維影像資訊區塊中的後擷取處理區塊可包含正規化欄位。
1 and 2 1 和 2
0
Y
Y
P147
Annex A 附錄 A
A post-acquisition processing block within a 3D image information block may contain unknown extensions. 3D 影像資訊區塊中的後處理區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P148
Annex A 附錄 A
A 3D image information block may contain a 3D textured image resolution block. 3D 影像資訊區塊可能包含 3D 紋理影像解析度區塊。
1 and 2 1 和 2
0
Y
Y
P149
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D mm shape x resolution field. 一個 3D 紋理影像解析度區塊可能包含一個 3D 毫米形狀 x 解析度欄位。
1 and 2 1 和 2
0
Y
Y
P150
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D mm shape y resolution field. 一個 3D 紋理影像解析度區塊可能包含一個 3D 毫米形狀 Y 解析度欄位。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P136 Annex A A post-acquisition processing block within a 3D image information block may contain a cropped field. 1 and 2 0 Y Y
P137 Annex A A post-acquisition processing block within a 3D image information block may contain a down-sampled field. 1 and 2 0 Y Y
P138 Annex A A post-acquisition processing block within a 3D image information block may contain a white-balance-adjusted field. 1 and 2 0 Y Y
P139 Annex A A post-acquisition processing block within a 3D image information block may contain a multiply-compressed field. 1 and 2 0 Y Y
P140 Annex A A post-acquisition processing block within a 3D image information block may contain an interpolated field. 1 and 2 0 Y Y
P141 Annex A A post-acquisition processing block within a 3D image information block may contain a contrast-stretched field. 1 and 2 0 Y Y
P142 Annex A A post-acquisition processing block within a 3D image information block may contain a pose-corrected field. 1 and 2 0 Y Y
P143 Annex A A post-acquisition processing block within a 3D image information block may contain a multi-view image field. 1 and 2 0 Y Y
P144 Annex A A post-acquisition processing block within a 3D image information block may contain an age-progressed field. 1 and 2 0 Y Y
P145 Annex A A post-acquisition processing block within a 3D image information block may contain a super-resolution processed field. 1 and 2 0 Y Y
P146 Annex A A post-acquisition processing block within a 3D image information block may contain a normalised field. 1 and 2 0 Y Y
P147 Annex A A post-acquisition processing block within a 3D image information block may contain unknown extensions. 1 and 2 0 Y Y
P148 Annex A A 3D image information block may contain a 3D textured image resolution block. 1 and 2 0 Y Y
P149 Annex A A 3D textured image resolution block may contain a 3D mm shape x resolution field. 1 and 2 0 Y Y
P150 Annex A A 3D textured image resolution block may contain a 3D mm shape y resolution field. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P136 | Annex A | A post-acquisition processing block within a 3D image information block may contain a cropped field. | 1 and 2 | 0 | Y | Y | | | |
| P137 | Annex A | A post-acquisition processing block within a 3D image information block may contain a down-sampled field. | 1 and 2 | 0 | Y | Y | | | |
| P138 | Annex A | A post-acquisition processing block within a 3D image information block may contain a white-balance-adjusted field. | 1 and 2 | 0 | Y | Y | | | |
| P139 | Annex A | A post-acquisition processing block within a 3D image information block may contain a multiply-compressed field. | 1 and 2 | 0 | Y | Y | | | |
| P140 | Annex A | A post-acquisition processing block within a 3D image information block may contain an interpolated field. | 1 and 2 | 0 | Y | Y | | | |
| P141 | Annex A | A post-acquisition processing block within a 3D image information block may contain a contrast-stretched field. | 1 and 2 | 0 | Y | Y | | | |
| P142 | Annex A | A post-acquisition processing block within a 3D image information block may contain a pose-corrected field. | 1 and 2 | 0 | Y | Y | | | |
| P143 | Annex A | A post-acquisition processing block within a 3D image information block may contain a multi-view image field. | 1 and 2 | 0 | Y | Y | | | |
| P144 | Annex A | A post-acquisition processing block within a 3D image information block may contain an age-progressed field. | 1 and 2 | 0 | Y | Y | | | |
| P145 | Annex A | A post-acquisition processing block within a 3D image information block may contain a super-resolution processed field. | 1 and 2 | 0 | Y | Y | | | |
| P146 | Annex A | A post-acquisition processing block within a 3D image information block may contain a normalised field. | 1 and 2 | 0 | Y | Y | | | |
| P147 | Annex A | A post-acquisition processing block within a 3D image information block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P148 | Annex A | A 3D image information block may contain a 3D textured image resolution block. | 1 and 2 | 0 | Y | Y | | | |
| P149 | Annex A | A 3D textured image resolution block may contain a 3D mm shape x resolution field. | 1 and 2 | 0 | Y | Y | | | |
| P150 | Annex A | A 3D textured image resolution block may contain a 3D mm shape y resolution field. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P151
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D mm shape z resolution field. 一個 3D 紋理影像解析度區塊可能包含一個 3D 毫米形狀 z 解析度欄位。
1 and 2 1 和 2
0
Y
Y
P152
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D mm texture resolution field. 一個 3D 紋理影像解析度區塊可能包含一個 3D 毫米紋理解析度欄位。
1 and 2 1 和 2
0
Y
Y
P153
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D texture acquisition period field. 一個 3D 紋理影像解析度區塊可能包含一個 3D 紋理擷取週期欄位。
1 and 2 1 和 2
0
Y
Y
P154
Annex A 附錄 A
A 3D textured image resolution block may contain a 3D face area scanned block. 一個 3D 紋理影像解析度區塊可能包含一個 3D 臉部區域掃描區塊。
1 and 2 1 和 2
0
Y
Y
P155
Annex A 附錄 A
A 3D face area scanned block may contain a front-of-the-head field. 一個 3D 臉部區域掃描區塊可能包含頭部前方區域。
1 and 2 1 和 2
0
Y
Y
P156
Annex A 附錄 A
A 3D face area scanned block may contain a chin field. 3D 臉部區域掃描區塊可能包含下巴欄位。
1 and 2 1 和 2
0
Y
Y
P157
Annex A 附錄 A
A 3D face area scanned block may contain an ears field. 3D 臉部區域掃描區塊可能包含耳朵欄位。
1 and 2 1 和 2
0
Y
Y
P158
Annex A 附錄 A
A 3D face area scanned block may contain a neck field. 3D 臉部區域掃描區塊可能包含頸部欄位。
1 and 2 1 和 2
0
Y
Y
P159
Annex A 附錄 A
A 3D face area scanned block may contain a back-of-the-head field. 3D 臉部區域掃描區塊可能包含後腦勺欄位。
1 and 2 1 和 2
0
Y
Y
P160
Annex A 附錄 A
A 3D face area scanned block may contain a full-head field. 3D 臉部區域掃描區塊可能包含全頭場域。
1 and 2 1 和 2
0
Y
Y
P161
Annex A 附錄 A
A 3D face area scanned block may contain unknown extensions. 3D 臉部區域掃描區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P162
Annex A 附錄 A
A 3D textured image resolution block may contain unknown extensions. 3D 紋理影像解析度區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P163
Annex A 附錄 A
A 3D image information block may contain a 3D texture map block. 3D 影像資訊區塊可能包含 3D 紋理貼圖區塊。
1 and 2 1 和 2
0
Y
Y
P164
Annex A 附錄 A
A 3D texture map block may contain an image data format field. 3D 紋理貼圖區塊可包含影像資料格式欄位。
1 and 2 1 和 2
0
Y
Y
P165
Annex A 附錄 A
A 3D texture map block may contain a 3D texture capture device spectral block spectrum field. 3D 紋理貼圖區塊可包含 3D 紋理擷取裝置光譜區塊光譜欄位。
1 and 2 1 和 2
0
Y
Y
P166
Annex A 附錄 A
A 3D texture map block may contain a texture 3D texture standard illuminant field. 3D 紋理貼圖區塊可包含紋理 3D 紋理標準光源欄位。
1 and 2 1 和 2
0
Y
Y
P167
Annex A 附錄 A
A 3D texture map block may contain a 3D error map field. 3D 紋理貼圖區塊可包含 3D 錯誤貼圖欄位。
1 and 2 1 和 2
0
Y
Y
P168
Annex A 附錄 A
A 3D texture map block may contain unknown extensions. 3D 紋理貼圖區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P169
Annex A 附錄 A
A 3D image information block may contain unknown extensions. 3D 影像資訊區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
P170
Annex A 附錄 A
A 3D shape representation block may contain unknown extensions. 3D 形狀表示區塊可能包含未知延伸。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P151 Annex A A 3D textured image resolution block may contain a 3D mm shape z resolution field. 1 and 2 0 Y Y
P152 Annex A A 3D textured image resolution block may contain a 3D mm texture resolution field. 1 and 2 0 Y Y
P153 Annex A A 3D textured image resolution block may contain a 3D texture acquisition period field. 1 and 2 0 Y Y
P154 Annex A A 3D textured image resolution block may contain a 3D face area scanned block. 1 and 2 0 Y Y
P155 Annex A A 3D face area scanned block may contain a front-of-the-head field. 1 and 2 0 Y Y
P156 Annex A A 3D face area scanned block may contain a chin field. 1 and 2 0 Y Y
P157 Annex A A 3D face area scanned block may contain an ears field. 1 and 2 0 Y Y
P158 Annex A A 3D face area scanned block may contain a neck field. 1 and 2 0 Y Y
P159 Annex A A 3D face area scanned block may contain a back-of-the-head field. 1 and 2 0 Y Y
P160 Annex A A 3D face area scanned block may contain a full-head field. 1 and 2 0 Y Y
P161 Annex A A 3D face area scanned block may contain unknown extensions. 1 and 2 0 Y Y
P162 Annex A A 3D textured image resolution block may contain unknown extensions. 1 and 2 0 Y Y
P163 Annex A A 3D image information block may contain a 3D texture map block. 1 and 2 0 Y Y
P164 Annex A A 3D texture map block may contain an image data format field. 1 and 2 0 Y Y
P165 Annex A A 3D texture map block may contain a 3D texture capture device spectral block spectrum field. 1 and 2 0 Y Y
P166 Annex A A 3D texture map block may contain a texture 3D texture standard illuminant field. 1 and 2 0 Y Y
P167 Annex A A 3D texture map block may contain a 3D error map field. 1 and 2 0 Y Y
P168 Annex A A 3D texture map block may contain unknown extensions. 1 and 2 0 Y Y
P169 Annex A A 3D image information block may contain unknown extensions. 1 and 2 0 Y Y
P170 Annex A A 3D shape representation block may contain unknown extensions. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P151 | Annex A | A 3D textured image resolution block may contain a 3D mm shape z resolution field. | 1 and 2 | 0 | Y | Y | | | |
| P152 | Annex A | A 3D textured image resolution block may contain a 3D mm texture resolution field. | 1 and 2 | 0 | Y | Y | | | |
| P153 | Annex A | A 3D textured image resolution block may contain a 3D texture acquisition period field. | 1 and 2 | 0 | Y | Y | | | |
| P154 | Annex A | A 3D textured image resolution block may contain a 3D face area scanned block. | 1 and 2 | 0 | Y | Y | | | |
| P155 | Annex A | A 3D face area scanned block may contain a front-of-the-head field. | 1 and 2 | 0 | Y | Y | | | |
| P156 | Annex A | A 3D face area scanned block may contain a chin field. | 1 and 2 | 0 | Y | Y | | | |
| P157 | Annex A | A 3D face area scanned block may contain an ears field. | 1 and 2 | 0 | Y | Y | | | |
| P158 | Annex A | A 3D face area scanned block may contain a neck field. | 1 and 2 | 0 | Y | Y | | | |
| P159 | Annex A | A 3D face area scanned block may contain a back-of-the-head field. | 1 and 2 | 0 | Y | Y | | | |
| P160 | Annex A | A 3D face area scanned block may contain a full-head field. | 1 and 2 | 0 | Y | Y | | | |
| P161 | Annex A | A 3D face area scanned block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P162 | Annex A | A 3D textured image resolution block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P163 | Annex A | A 3D image information block may contain a 3D texture map block. | 1 and 2 | 0 | Y | Y | | | |
| P164 | Annex A | A 3D texture map block may contain an image data format field. | 1 and 2 | 0 | Y | Y | | | |
| P165 | Annex A | A 3D texture map block may contain a 3D texture capture device spectral block spectrum field. | 1 and 2 | 0 | Y | Y | | | |
| P166 | Annex A | A 3D texture map block may contain a texture 3D texture standard illuminant field. | 1 and 2 | 0 | Y | Y | | | |
| P167 | Annex A | A 3D texture map block may contain a 3D error map field. | 1 and 2 | 0 | Y | Y | | | |
| P168 | Annex A | A 3D texture map block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P169 | Annex A | A 3D image information block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P170 | Annex A | A 3D shape representation block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
Table C. 2 (continued) 表 C. 2 (續)
Provision identifier 條款識別碼
Reference in data format specification 資料格式規範中的參考
Provision summary 條文摘要
Level 等級
Status 狀態
Format type applicability 格式類型適用性
IUT support IUT 支援
Supported range 支援範圍
Test result 檢查結果
P171
Annex A 附錄 A
A 3D representation kind block may contain a 3D vertex block. 3D 表示種類區塊可包含 3D 頂點區塊。
1 and 2 1 和 2
0
Y
Y
P172
Annex A 附錄 A
A 3D vertex block may contain a 3D vertex information block. 3D 頂點區塊可包含 3D 頂點資訊區塊。
1 and 2 1 和 2
0
P173
AnnexA 附錄 A
A 3D vertex information block may contain a 3D vertex identifier. 3D 頂點資訊區塊可包含 3D 頂點識別碼。
1 and 2 1 和 2
0
Y
Y
P174
Annex A 附錄 A
A3D vertex information block may contain a 3D vertex normals block. A3D 頂點資訊區塊可能包含 3D 頂點法線區塊。
1 and 2 1 和 2
0
Y
Y
P175
Annex A 附錄 A
A 3D vertex information block may contain a 3D textures block. 3D 頂點資訊區塊可能包含 3D 紋理區塊。
1 and 2 1 和 2
0
Y
Y
P176
Annex A 附錄 A
A 3D vertex information block may contain a 3D error map field. 3D 頂點資訊區塊可能包含 3D 錯誤對應欄位。
1 and 2 1 和 2
0
Y
Y
P177
Annex A 附錄 A
A 3D vertex information block may contain unknown extensions. 3D 頂點資訊區塊可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
P178
Annex A 附錄 A
A 3D vertex block may contain a 3D vertex triangle data block. 3D 頂點區塊可包含 3D 頂點三角形資料區塊。
1 and 2 1 和 2
0
Y
Y
P179
Annex A 附錄 A
A 3D representation kind may contain unknown extensions. 3D 表示類型可能包含未知的擴充功能。
1 and 2 1 和 2
0
Y
Y
Provision identifier Reference in data format specification Provision summary Level Status Format type applicability IUT support Supported range Test result
P171 Annex A A 3D representation kind block may contain a 3D vertex block. 1 and 2 0 Y Y
P172 Annex A A 3D vertex block may contain a 3D vertex information block. 1 and 2 0
P173 AnnexA A 3D vertex information block may contain a 3D vertex identifier. 1 and 2 0 Y Y
P174 Annex A A3D vertex information block may contain a 3D vertex normals block. 1 and 2 0 Y Y
P175 Annex A A 3D vertex information block may contain a 3D textures block. 1 and 2 0 Y Y
P176 Annex A A 3D vertex information block may contain a 3D error map field. 1 and 2 0 Y Y
P177 Annex A A 3D vertex information block may contain unknown extensions. 1 and 2 0 Y Y
P178 Annex A A 3D vertex block may contain a 3D vertex triangle data block. 1 and 2 0 Y Y
P179 Annex A A 3D representation kind may contain unknown extensions. 1 and 2 0 Y Y | Provision identifier | Reference in data format specification | Provision summary | Level | Status | Format type applicability | | IUT support | Supported range | Test result |
| :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- | :--- |
| P171 | Annex A | A 3D representation kind block may contain a 3D vertex block. | 1 and 2 | 0 | Y | Y | | | |
| P172 | Annex A | A 3D vertex block may contain a 3D vertex information block. | 1 and 2 | 0 | | | | | |
| P173 | AnnexA | A 3D vertex information block may contain a 3D vertex identifier. | 1 and 2 | 0 | Y | Y | | | |
| P174 | Annex A | A3D vertex information block may contain a 3D vertex normals block. | 1 and 2 | 0 | Y | Y | | | |
| P175 | Annex A | A 3D vertex information block may contain a 3D textures block. | 1 and 2 | 0 | Y | Y | | | |
| P176 | Annex A | A 3D vertex information block may contain a 3D error map field. | 1 and 2 | 0 | Y | Y | | | |
| P177 | Annex A | A 3D vertex information block may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
| P178 | Annex A | A 3D vertex block may contain a 3D vertex triangle data block. | 1 and 2 | 0 | Y | Y | | | |
| P179 | Annex A | A 3D representation kind may contain unknown extensions. | 1 and 2 | 0 | Y | Y | | | |
ISO/IEC 39794-5:2019(E)
IUT support notes IUT 支援注意事項
To be filled in by the supplier of the IUT on the copy of this table provided to the testing laboratory and to be included in the copy of this table that forms part of the test report. 由 IUT 供應商填寫於提供給測試實驗室的本表副本中,並應包含於構成測試報告一部分的本表副本中。
Test result notes 測試結果備註
To be filled in by the testing laboratory, if necessary, during the execution of the conformance test and to be included in the copy of this table that forms part of the test report. 由測試實驗室在執行一致性測試期間(如有必要)填寫,並應包含於構成測試報告一部分的本表副本中。
C. 3 Conformance test assertions C.3 一致性測試斷言
Level 1 and 2 requirements and options shall be tested by: 等級 1 和等級 2 的要求與選項應透過以下方式進行測試:
decoding tagged binary data blocks under test based on the ASN. 1 module that specifies the tagged binary data format; or 根據指定標記二進位資料格式的 ASN.1 模組,解碼受測的標記二進位資料區塊;或
validation of XML documents under test against the XML schema definition that specifies the textual data format, respectively. 分別根據指定文字資料格式的 XML 綱要定義,驗證受測的 XML 文件。
C. 4 Conformance testing for profiles given in Annex D C.4 附錄 D 中所列規範的符合性測試
This clause specifies conformance testing methodologies for the specific requirements according to the application profiles as given in Annex D. 本條款規定了根據附錄 D 中給出的應用規範,針對特定要求的符合性測試方法。
NOTE Currently, no conformance testing methodologies for the application profiles in Annex D are available for this document. 註:目前,本文件尚無適用於附錄 D 中應用規範的符合性測試方法。
Annex D (normative) 附錄 D (規範性)
Application profiles 應用設定檔
D. 1 Reference face image for Machine Readable Travel Documents (MRTD) D.1 機器可讀旅行證件 (MRTD) 的參考臉部影像
D.1.1 General D.1.1 概述
ICAO Doc 9303 provides the basic functional specification for MRTDs and describes all relevant properties of MRTDs. ICAO Doc 9303 提供了 MRTDs 的基本功能規範,並描述了 MRTDs 的所有相關屬性。
The face portrait printed on the ICAO compliant MRTD is an essential element of that document and one of the most important information carriers binding the document to the holder. A standardized face portrait produced at a high quality helps issuing agencies to screen identity and border agencies to inspect the travel document manually or via automated processing. 印製在符合 ICAO 規範的 MRTD 上的臉部肖像,是該文件的一個重要元素,也是將文件與持有人綁定的最重要資訊載體之一。高品質的標準化臉部肖像有助於發證機構篩選身分,並協助邊境機構手動或透過自動化處理檢查旅行文件。
After the introduction of the digitally stored image in 2005, ABC (Automated Border Control) systems have been introduced to perform automated comparison of the individual and the electronically stored image. Those ABCA B C systems compare, whether it is manually or automated, the printed image and/or the electronically stored image and the image taken live while crossing a border. 在 2005 年引入數位儲存影像後,ABC(自動邊境管制)系統已被引入,用於執行個人與電子儲存影像的自動比較。這些系統無論是手動還是自動,都會比較列印影像和/或電子儲存影像,以及過境時即時拍攝的影像。
This annex contains significant content from Reference [38]. 本附錄包含大量來自參考文獻[38]的內容。
D.1.2 Overview D.1.2 概述
This annex describes the requirements and best practice recommendations to be applied for face portrait capturing in the application case of enrolment of biometric reference data for electronic MRTD. In this sense, this annex is an application profile. 本附錄說明了在電子機器可讀旅行證件(MRTD)中註冊生物識別參考資料的應用案例中,人臉肖像擷取所需遵循的要求和最佳實踐建議。因此,本附錄是一個應用規範。
This annex: 本附錄:
shares the lessons learned using the stored and displayed face portrait in an MRTD, 分享了使用儲存並顯示於 MRTD 中的人臉肖像所學到的經驗,
describes how the face portraits should be captured that serve as the content of this document and its data structures, 說明了應如何擷取人臉肖像,以作為本文件及其資料結構的內容,
provides the experiences made applying face recognition technology in ABC gates, manual border control, identity screening, and other applications based on the face portraits provided by electronic MRTDs. It also gives guidance on the requirements for capturing and processing face portraits contained in MRTDs to support the inspection process, and 提供在 ABC 閘門、人工邊境管制、身分查驗以及其他基於電子機器可讀旅行證件(MRTD)所提供之臉部肖像的應用中,實施臉部辨識技術的經驗。它也針對擷取和處理 MRTD 中包含的臉部肖像以支援查驗過程的需求,提供指導,並且
provides comprehensive recommendations for face portrait capturing including scene, photographic and digital requirements. 為臉部肖像的擷取提供全面的建議,包括場景、攝影和數位要求。
The following topics are not in scope of this annex, requirements on them are given in ICAO Doc 9303: 以下主題不屬於本附錄的範圍,其要求已在 ICAO Doc 9303 中說明:
image printing and scanning as well as on digital image processing, 影像列印和掃描以及數位影像處理,
face portraits printed on MRTDs to ensure good visibility for inspection, 用於 MRTD 上的人臉肖像,以確保檢查時的良好可見度,
guidance for reader system manufacturers on the use of unified reflection free illumination and view angles, and 為讀取器系統製造商提供關於使用統一無反射照明和視角的指導,以及
image capturing for verification and/or identification applications like ABC , even if many of the requirements listed in this document apply to such images, too. 用於驗證和/或識別應用(如 ABC)的影像擷取,即使本文件中列出的許多要求也適用於此類影像。
ISO/IEC 39794-5:2019(E)
The following topics are not in scope of this annex: 本附錄不涵蓋以下主題:
Definition of image data formats like JPEG, JPEG2000, PNG, 影像資料格式的定義,例如 JPEG、JPEG2000、PNG,
Security aspects like digital image electronic signature, presentation attack detection (PAD), and morphing prevention. 安全性方面,例如數位影像電子簽章、呈現攻擊偵測(PAD)和變臉預防。
For certain criteria, there may be two different levels given in a table form: A minimum requirement and a best practice recommendation. The requirement gives the minimum acceptable values or value ranges in order to reach compliance. The best practice recommendation gives values that will result in better overall performance or quality, and users are encouraged to adopt best practice values whenever possible. See Table D.1. 對於某些標準,可能會以表格形式提供兩種不同等級:最低要求和最佳實踐建議。最低要求提供可接受的最小值或數值範圍,以達到合規性。最佳實踐建議則提供能帶來更好整體效能或品質的數值,鼓勵使用者盡可能採用最佳實踐值。請參閱表 D.1。
Table D. 1 - Sample table summarizing minimum requirements and best practice recommendations 表 D.1 - 彙總最低要求和最佳實踐建議的範例表格
Criterion 標準
Requirement 要求
The criterion shall be ... 該標準應為...
Best practice 最佳實務
The criterion should be ... 該標準應為…
Criterion Requirement The criterion shall be ...
Best practice The criterion should be ...| Criterion | Requirement | The criterion shall be ... |
| :--- | :--- | :--- |
| | Best practice | The criterion should be ... |
D.1.3 Face portraits D.1.3 臉部肖像
D.1.3.1 Uses of face portrait images D.1.3.1 臉部肖像影像的用途
Face portraits appear in several places on and in an MRTD: 臉部肖像出現在 MRTD 的數個位置:
As a printed image on the data page (Zone V as defined in ICAO Doc 9303), 作為資料頁上的列印圖像(國際民航組織第 9303 號文件定義的 V 區),
As a digital image stored in the RFID chip, 作為儲存在 RFID 晶片中的數位圖像,
Optionally, as a secondary image on the data and/or observation page, e.g., as a changeable laser image, as a micro-perforation, or as a background print. 可選擇性地作為資料及/或觀察頁面上的次要影像,例如作為可變雷射影像、微穿孔或背景列印。
All the images used shall be derived from the same captured face portrait. However, the technical requirements of each of the images may differ depending on the applied technology. 所有使用的影像應源自同一張擷取的人臉肖像。然而,每張影像的技術要求可能因所應用的技術而異。
The intended use for a printed face portrait is to give a good physical representation of the document holder and to allow for a human comparison of the face portrait and the holder of the MRTD. Physical security features and the printing technology may interact or influence the face portrait which needs to be considered as part of the comparison process. 列印人臉肖像的預期用途是提供文件持有人的良好實體表徵,並允許人為比較人臉肖像與 MRTD 持有人。實體安全功能和列印技術可能會相互作用或影響人臉肖像,這需要作為比較過程的一部分加以考慮。
The intended use of the face portrait image digitally stored in the chip is such that the image can be compared to the printed face portrait and the human via manual processes or compared to a live image via automated processes in a 1:1 od 1:N application case. Because of the way the image is stored on the MRTD (see Doc 9303 Part 11), border agencies can confirm that the image has been stored on the MRTD by the issuing authority and remains unaltered or unsubstituted. The digital image is the primary image used for biometric comparison. 晶片中數位儲存的人臉肖像影像的預期用途是,該影像可透過人工程序與列印人臉肖像和人進行比較,或透過自動程序在 1:1 或 1:N 應用案例中與即時影像進行比較。由於影像儲存在 MRTD 上的方式(參見 Doc 9303 第 11 部分),邊境機構可以確認影像已由發證機關儲存在 MRTD 上,並且未經更改或替換。數位影像是用於生物識別比較的主要影像。
The secondary images serve as physical security features protecting the printed face portrait. Therefore their appearance shall be the same as the printed face portrait. However, size and production technology determine the technical requirements of the face portrait derivative used here. 次要影像作為實體安全功能,保護列印的人臉肖像。因此,它們的外觀應與列印的人臉肖像相同。然而,尺寸和生產技術決定了此處使用的人臉肖像衍生性商品之技術要求。
Figure D. 1 displays the appearance of the face portrait as a printed image and as a digitally stored image. 圖 D.1 顯示了人臉肖像作為列印影像和數位儲存影像的外觀。
Figure D. 1 - Face portraits to be used on/in an MRTD 圖 D.1 - 用於 MRTD 上/內的人臉肖像
Two main types of application processes are considered, those based on: 主要考慮兩種應用流程類型,其基礎為:
submission of printed photographs provided by the citizen to the passport authority, and on 公民向護照主管機關提交紙本照片,以及
electronic face portrait submission. 電子人像提交。
There are two sub-types of the second type: 第二種類型有兩種子類型:
live capture where the applicant has the photo taken during an interview or application submission, and 即時拍攝,申請人在面試或提交申請時拍攝照片,以及
upload, where the image is provided electronically by the applicant, by an enrolment bureau or by an accredited ID photo service. 上傳,其中影像由申請人、註冊局或經認可的身份證照片服務單位以電子方式提供。
These two sub-types are subject to the same requirements. Depending on the process type, different clauses of this document apply, as defined below. Both main types are shown in Figure D.2. 這兩種子類型受相同要求約束。根據處理類型,本文件的不同條款適用,如下所述。兩種主要類型如圖 D.2 所示。
The production of the printed as well as of the electronic face portraits may be done by automated photo kiosks, officers of passport authorities, photo booths or photographers. It is essential that the quality requirements are met. Photographic experts should be consulted before introducing a new enrolment solution. 列印及電子人像照片的製作可透過自動照相亭、護照主管機關人員、照相亭或攝影師完成。重要的是必須符合品質要求。在引入新的註冊解決方案之前,應諮詢攝影專家。
Figure D. 2 - Face portrait enrolment process variations 圖 D. 2 - 人像註冊流程變體
D.1.3.2 Passport application using printed face portraits D.1.3.2 使用列印人像照片的護照申請
For a passport application process that uses printed face portraits the citizen typically visits a photographer or photo booth to obtain such a face portrait. In all cases the citizen receives printed photos, and there is no electronic submission or linkage to an electronically stored image available. Then the citizen submits such a photo to the passport issuing authority as part of their application. To establish a passport application process using printed face portraits, D.1.4 and D.1.5 apply. Additional requirements on image printing for submission purposes, scanning of printed images and image printing on MRTD data pages specified in ICAO Doc 9303 apply. 對於使用列印人像照片的護照申請流程,公民通常會造訪攝影師或照相亭以取得此類人像照片。在所有情況下,公民都會收到列印的照片,並且沒有電子提交或連結到可用的電子儲存影像。然後,公民將此類照片作為其申請的一部分提交給護照核發機關。為建立使用列印人像照片的護照申請流程,適用 D.1.4 和 D.1.5。國際民航組織 Doc 9303 中規定的用於提交目的的影像列印、列印影像的掃描以及 MRTD 資料頁面上的影像列印的額外要求亦適用。
D.1.3.3 Passport application using electronic face portrait submission D.1.3.3 使用電子人像提交的護照申請
Enrolment data providers, such as photographers, photo booths or kiosks can be linked electronically to the issuing authorities. For a passport application that uses electronic submission, the intermediate steps dealing with a printed photo are skipped. In most cases, the photo is digitally captured and electronically stored or directly transmitted to the passport issuing authority. There are many ways the face portrait may be transferred to the passport issuing authority. Such schemas include direct transmission to the authority, a data carrier submitted by the citizen, and temporary storage on a server and submission of a reference to the uploaded/stored photo provided by the citizen. Live capture where the applicant has the photo taken during an interview or application submission is covered here, too. To establish a passport application process using electronic face portraits, D.1.4 and D.1.5 apply. Additional requirements on image printing on MRTD data pages specified in ICAO Doc 9303 apply. The IED requirement for electronically submitted images follows the same requirement for chip images as in D.1.5 and Tables D. 4 and D.10. 註冊資料提供者,例如攝影師、照相亭或自助服務機,可以電子方式連結至發證機關。對於使用電子提交的護照申請,可跳過處理列印照片的中間步驟。在大多數情況下,照片會以數位方式擷取並電子儲存,或直接傳輸至護照發證機關。人像可透過多種方式傳輸至護照發證機關。此類方案包括直接傳輸至機關、公民提交的資料載體,以及在伺服器上暫時儲存並提交公民提供的上傳/儲存照片的參考。申請人在面談或提交申請期間拍攝照片的即時擷取也涵蓋於此。為建立使用電子人像的護照申請流程,適用 D.1.4 和 D.1.5。國際民航組織文件 9303 中規定的關於 MRTD 資料頁面影像列印的額外要求也適用。電子提交影像的 IED 要求與 D.1.5 和表 D.4 和 D.10 中晶片影像的要求相同。
Passport applicants should not be encouraged to submit face portraits captured by amateur photographers or captured on amateur equipment such as mobile phones or tablets or printed on consumer printers (home-made face portraits) as they typically do not achieve the required quality level as specified in D.1.4. If an issuer decides to accept homemade face portraits, the issuer shall 不應鼓勵護照申請人提交由業餘攝影師拍攝或使用業餘設備(例如行動裝置或平板電腦)拍攝,或在消費型印表機(自製人像)上列印的人像,因為它們通常無法達到 D.1.4 中規定的所需品質水準。如果發證機關決定接受自製人像,則發證機關應
ensure, based on an appropriate level of expertise, that the printing quality and all of the requirements specified in D.1.4 are maintained, and that the risks of photo manipulation and morphing inherent with such an uncontrolled process are suitably mitigated. 確保在適當的專業知識水準下,維持列印品質及 D.1.4 中規定的所有要求,並適當降低此類不受控流程所固有的照片篡改和變臉風險。
D.1.4 Enrolment live face portrait capturing D.1.4 註冊即時臉部肖像擷取
D.1.4.1 General D.1.4.1 一般
This clause describes the requirements for the environment that is used for face portrait capturing. Additionally, it gives recommendations on best practice. The requirements on the environment are derived from experiences made in face recognition applications including ABC gates, and they consider the methods used by professional photographers. 本條款說明用於人臉肖像擷取的環境要求。此外,它還提供了最佳實踐建議。環境要求源自於人臉辨識應用(包括 ABC 閘門)的經驗,並考量了專業攝影師所使用的方法。
Before introducing new equipment and defining processes for enrolment data capturing, an experienced face portrait photographer and/or an optics expert should be asked for advice. The requirements apply to all installations including photo booths and kiosks. 在引進新設備和定義註冊資料擷取流程之前,應諮詢經驗豐富的人像攝影師和/或光學專家以獲取建議。這些要求適用於所有裝置,包括照相亭和自助服務機。
This clause specifies requirements for the photograph being captured as well as for the photographic equipment being used. Figure D. 3 shows the content of D.1.4 in the MRTD production process chain. 本條款規定了所拍攝照片以及所用攝影設備的要求。圖 D.3 顯示了 MRTD 生產流程鏈中 D.1.4 的內容。
Figure D. 3 - Content of D.1.4 in the MRTD production process chain (boxed in red) 圖 D.3 - MRTD 生產流程鏈中 D.1.4 的內容(紅框標示)
NOTE Pose constraints are very difficult to evaluate on the acquired 2D image, even for experts in this field. Numeric values have been provided in this document to support a consistent subject positioning in the fullfrontal pose. 註:即使對於該領域的專家而言,也很難評估所獲取 2D 影像中的姿勢限制。本文件提供了數值,以支援在完全正面姿勢下保持一致的主體定位。
D.1.4.2 Camera and scene D.1.4.2 相機與場景
D.1.4.2.1 Selection of camera and focal length D.1.4.2.1 攝影機與焦距的選擇
In addition to choosing an appropriate camera-to-subject distance (CSD), as described in D.1.4.2.2, the selection of a camera and its lens is a major factor affecting the quality of face portrait images. To ensure 除了選擇適當的相機到拍攝對象距離(CSD),如 D.1.4.2.2 所述,相機及其鏡頭的選擇是影響臉部肖像影像品質的主要因素。為確保
high image quality and a standards-compliant inter-eye distance (IED), the camera’s sensor must have sufficient pixel dimensions and its lens must be chosen to match its image sensor’s physical dimensions. 高影像品質和符合標準的雙眼距離(IED),相機的感測器必須具有足夠的像素尺寸,並且其鏡頭必須選擇與其影像感測器的物理尺寸相匹配。
For example, for a camera using an APS-C sensor (having a crop factor of 1,44), photographers should consider using a lens of focal length between 50/1,44 and 130/1,44, or roughly 35 mm to 90 mm . 例如,對於使用 APS-C 感測器(裁切係數為 1.44)的相機,攝影師應考慮使用焦距介於 50/1.44 和 130/1.44 之間,或大約 35 毫米至 90 毫米的鏡頭。
For face portrait photos, photographers using a conventional 35 mm film camera (having a 36mmxx24mm36 \mathrm{~mm} \times 24 \mathrm{~mm} frame, with a 43,3mm43,3 \mathrm{~mm} diagonal) often select a normal to moderate telephoto lens, with a focal length between 50 and 130 mm (or an equivalent zoom lens). For digital cameras employing typically smaller size CMOS or CCD image sensors, the lens selected for face portrait photography should have a proportionally decreased focal length. 對於臉部肖像照片,攝影師使用傳統的 35 毫米底片相機(具有 36mmxx24mm36 \mathrm{~mm} \times 24 \mathrm{~mm} 畫幅,對角線為 43,3mm43,3 \mathrm{~mm} )通常會選擇標準到中等望遠鏡頭,焦距介於 50 到 130 毫米之間(或等效的變焦鏡頭)。對於採用通常較小尺寸 CMOS 或 CCD 影像感測器的數位相機,用於臉部肖像攝影的鏡頭應具有按比例縮小的焦距。
For further explanations on sensor diagonal and sensor diagonal encoding see 7.42 _\underline{7.42} and 7.43 _\underline{7.43}. 有關感測器對角線和感測器對角線編碼的進一步說明,請參閱 7.42 _\underline{7.42} 和 7.43 _\underline{7.43} 。
Figure D. 4 illustrates the typical optical arrangement and terminology for face portrait image acquisition, as well as some of the variables in the arrangement. 圖 D.4 闡述了臉部肖像影像擷取典型的光學配置與術語,以及配置中的一些變數。
Figure D.4 - Illustration of optical arrangement and terminology 圖 D.4 — 光學配置與術語說明
For a selected CSD (in millimetres), and camera image sensor with a vertical dimension in millimetres hh, a requested field of view of H_("FieldofViewmm ")H_{\text {FieldofViewmm }}, the focal length f (in millimetres) can be computed using the following relationships in order to optimise the requested field of view of the subject into the sensor dimensions: 對於選定的 CSD(單位為公釐),以及垂直尺寸為公釐的相機影像感測器 hh ,所要求的視野為 H_("FieldofViewmm ")H_{\text {FieldofViewmm }} ,焦距 f(單位為公釐)可使用以下關係式計算,以將主體的所需視野最佳化至感測器尺寸:
f~=h_(mm)(CSD)/(H_("FieldOfViewmm "))f \cong h_{m m} \frac{C S D}{H_{\text {FieldOfViewmm }}}
In case of homemade face portraits, the lens optimization is not done due to the large camera angle. 在自製臉部肖像的情況下,由於相機角度較大,因此未進行鏡頭最佳化。
For the same camera image sensor with vertical pixel count of hh pixels, the inter-eye distance on the sensor in pixels IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} may be computed using the following relation-ships, where IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} is intereye distance in millimetres on the subject. 對於垂直像素數為 hh 像素的相同相機影像感測器,感測器上像素間的眼距 IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} 可使用以下關係式計算,其中 IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} 為受測者上以毫米計的眼距。
IED_(mm)^("Sensor ")~=IED_(mm)^("Subject ")xx(f)/(CSD)I E D_{m m}^{\text {Sensor }} \cong I E D_{m m}^{\text {Subject }} \times \frac{f}{C S D}
and 和
IED_(px)^("Sensor ")=IED_(mm)^("Sensor ")xx(h_(px))/(h_(mm))I E D_{p x}^{\text {Sensor }}=I E D_{m m}^{\text {Sensor }} \times \frac{h_{p x}}{h_{m m}}
EXAMPLE 1 A commercially available, digital single lens reflex (DSLR) camera has the following specifications: sensor APS-C, 22,3mmxx14,9mm,5184pxxx3456px,1822,3 \mathrm{~mm} \times 14,9 \mathrm{~mm}, 5184 \mathrm{px} \times 3456 \mathrm{px}, 18 megapixels. For a CSD of 1200 mm , a typical H_("FieldofViewmm ")H_{\text {FieldofViewmm }} of 500 mm , a typical IED of the subject of about 62 mm the calculations below show that the focal length ff will be about 50 mm (equivalent to 80 mm full frame). 範例 1 市售的數位單眼反光(DSLR)相機具有以下規格:感測器 APS-C, 22,3mmxx14,9mm,5184pxxx3456px,1822,3 \mathrm{~mm} \times 14,9 \mathrm{~mm}, 5184 \mathrm{px} \times 3456 \mathrm{px}, 18 百萬像素。對於 1200 毫米的 CSD、典型的 500 毫米 H_("FieldofViewmm ")H_{\text {FieldofViewmm }} 、以及受測者約 62 毫米的典型 IED,以下計算顯示焦距 ff 約為 50 毫米(相當於 80 毫米全片幅)。
The calculations below show that IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} will be about 598 pixels, well above the best practice value suggested in Table D.4. 以下計算顯示 IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} 約為 598 像素,遠高於表 D.4 中建議的最佳實踐值。
IED_(px)^("Sensor ")~=2,58mmxx(3456px)/(14,9(mm))=598" pixels "I E D_{p x}^{\text {Sensor }} \cong 2,58 \mathrm{~mm} \times \frac{3456 \mathrm{px}}{14,9 \mathrm{~mm}}=598 \text { pixels }
EXAMPLE 2 For a sensor of 5 megapixels ( 2592pxxx1944px2592 \mathrm{px} \times 1944 \mathrm{px} ) with an optimized focal length lens, the IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} will be about 336 pixels, well above the best practice value suggested in Table D.4: 範例 2 對於一個 5 百萬像素( 2592pxxx1944px2592 \mathrm{px} \times 1944 \mathrm{px} )的感測器,搭配最佳化焦距的鏡頭,其 IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} 將約為 336 像素,遠高於表 D.4 中建議的最佳實踐值:
IED_(px)^("Sensor ")~=2,58mmxx(1944px)/(14,9(mm))=336" pixels "I E D_{p x}^{\text {Sensor }} \cong 2,58 \mathrm{~mm} \times \frac{1944 \mathrm{px}}{14,9 \mathrm{~mm}}=336 \text { pixels }
D.1.4.2.2 Magnification distortion and camera subject distance D.1.4.2.2 放大失真與相機主體距離
All images captured by a photographic system will contain image distortion. Every face portrait is a compromise between different requirements like camera and lens costs or available space and illumination. In this document requirements and recommendations are given to ensure global interoperability in the sense that the most important properties of every face portrait used for MRTD purposes reach the correct quality requirements and therefore ensure similar performance in face image-based authentication applications like border control systems. 所有由攝影系統拍攝的影像都將包含影像失真。每張人臉肖像都是在不同要求之間妥協的結果,例如相機和鏡頭成本,或可用空間和照明。本文件提供了要求和建議,以確保全球互通性,意即用於 MRTD 目的的每張人臉肖像最重要的屬性都能達到正確的品質要求,從而確保在邊境管制系統等基於人臉影像的驗證應用中具有相似的效能。
The CSD requirements are listed in Table D.2. For sample face portraits illustrating possible effects of the optical system see Figures D. 6 and D.7. Table D. 4 lists different camera subject distances and their corresponding magnification distortions. CSD 要求列於表 D.2。如需說明光學系統可能效果的臉部肖像範例,請參閱圖 D.6 和 D.7。表 D.4 列出不同的相機主體距離及其對應的放大失真。
Magnification distortion can only be evaluated by measuring tools. See E.2. It is not possible to evaluate magnification distortion value from human vision. For information, ears start to be masked around a magnification distortion of 14%14 \% or higher. 放大失真只能透過測量工具進行評估。請參閱 E.2。無法透過人眼評估放大失真值。供參考,當放大失真達到 14%14 \% 或更高時,耳朵開始被遮蔽。
NOTE 1 Selfie-style face portraits are likely not to maintain the minimal distance requirement. 註 1 自拍風格的臉部肖像可能無法維持最小距離要求。
Table D. 2 - CSD requirements and recommendations 表 D.2 - CSD 要求與建議
1,2m <= CSD <= 2,5m1,2 \mathrm{~m} \leq C S D \leq 2,5 \mathrm{~m}
Criterion: CSD for 1:1 Requirement 0,7m <= CSD <= 4m
Best practice 1,0m <= CSD <= 2,5m
Criterion: CSD for 1:N Requirement 1m <= CSD <= 4m
Best practice 1,2m <= CSD <= 2,5m| Criterion: CSD for 1:1 | Requirement | $0,7 \mathrm{~m} \leq \mathrm{CSD} \leq 4 \mathrm{~m}$ |
| :--- | :--- | :--- |
| | Best practice | $1,0 \mathrm{~m} \leq \mathrm{CSD} \leq 2,5 \mathrm{~m}$ |
| Criterion: CSD for 1:N | Requirement | $1 \mathrm{~m} \leq \mathrm{CSD} \leq 4 \mathrm{~m}$ |
| | Best practice | $1,2 \mathrm{~m} \leq C S D \leq 2,5 \mathrm{~m}$ |
The camera shall be at the subject’s eye-level. The line between camera and centre of subject’s face shall be horizontal within a maximum HD of +-5^(@)\pm 5^{\circ}. Height alignment should be done by vertical adjustment of either subject or camera. See Figure D.5. 攝影機應與拍攝對象的眼睛高度齊平。攝影機與拍攝對象臉部中心之間的連線應保持水平,最大水平偏差(HD)為 +-5^(@)\pm 5^{\circ} 。高度對齊應透過垂直調整拍攝對象或攝影機來完成。請參閱圖 D.5。
Figure D. 5 - Alignment of camera and subject 圖 D.5 - 攝影機與拍攝對象的對齊
These recommendations and requirements apply for all capturing setups including photo booths and kiosks. 這些建議和要求適用於所有拍攝設置,包括照相亭和自助服務機。
Figure D. 6 - Appearance with and without strong magnification distortion 圖 D. 6 - 有無強烈放大失真時的外觀
One of the important factors that influence the appearance of the facial features is the distance between subject and camera lens. 影響臉部特徵外觀的重要因素之一是受試者與相機鏡頭之間的距離。
The magnification distortion due to camera subject distance can be noticeable to human examiners but shall be within defined limits that allow effective face recognition. 由於相機與受試者距離造成的放大失真,人類檢查員可能會注意到,但其應在允許有效人臉辨識的定義限制內。
Acceptable distortion rate tolerances depend on the performance capacity of state of the art face recognition technology, and on the capability of typical human inspection staff to recognize people, even those coming from varying ethnic origin. 可接受的失真率容許值取決於最先進人臉辨識技術的效能,以及一般人工檢查人員辨識人臉的能力,即使是來自不同種族的人。
a) Distance of 300mm\mathbf{3 0 0 ~ m m} a) 距離為 300mm\mathbf{3 0 0 ~ m m}
b) Distance of 400mm\mathbf{4 0 0 ~ m m} b) 距離為 400mm\mathbf{4 0 0 ~ m m}
c) Distance of 600mm\mathbf{6 0 0 ~ m m} c) 距離為 600mm\mathbf{6 0 0 ~ m m}
Figure D. 7 - Sample face portraits taken with a full-size sensor camera at focal length 50 mm from different distances 圖 D. 7 - 使用全尺寸感測器相機在 50 mm 焦距下從不同距離拍攝的臉部肖像範例
These images have been captured using the enrolment bench described in E.6. All images have been normalized to a constant IED. The red bars mark the distance between the feature points 10.7 and 10.8 according to ISO/IEC 14496-2:2004, Annex C measured in Figure D. 7 i). 這些影像已使用 E.6 中所述的註冊台進行擷取。所有影像均已正規化為恆定 IED。紅色條標示了根據 ISO/IEC 14496-2:2004,附錄 C 在圖 D. 7 i) 中測量的特徵點 10.7 和 10.8 之間的距離。
Rulers at nose and ear may be used to measure the geometric effect to the face, i.e., a millimetre at nose level is larger than a millimetre at ear level on the image of the rulers. 鼻子和耳朵處的尺規可用於測量對臉部的幾何效應,即,在尺規影像上,鼻子水平處的一毫米大於耳朵水平處的一毫米。
The maximum level of magnification distortion of the capturing process shall be set depending on the appropriate application case: 擷取過程的最大放大失真程度應根據適當的應用案例設定:
1:1 application case: At the border, an automatic and/or human face verification/comparison is progressed. This is the case in most automated border control applications. The maximum magnification distortion rate of the picture in the passport shall not be greater than 7%7 \% and ideally should not be greater than 5 %. 1:1 應用案例:在邊境,進行自動和/或人工臉部驗證/比對。大多數自動邊境管制應用都屬於此類。護照中圖片的最大放大失真率不應大於 7%7 \% ,理想情況下不應大於 5%。
1:N application case: At the enrolment or issuance time of the document, a 1:N face identification is done on a database to help verifying the uniqueness of the identity associated to the new image provided. N is as large as the number of images searched. This application case requires higher quality enrolment. The maximum magnification distortion rate shall not be greater than 5%5 \% and ideally should not be greater than 4 %. 1:N 應用案例:在文件註冊或發行時,對資料庫進行 1:N 臉部識別,以協助驗證與所提供新影像相關聯的身份的唯一性。N 與搜尋的影像數量一樣大。此應用案例需要更高品質的註冊。最大放大失真率不應大於 5%5 \% ,理想情況下不應大於 4%。
The study presented in E. 6 has shown that, for a large range of enrolment and verification distances the influence of magnification distortion on automatic face recognition system performance is low. E.6 中提出的研究顯示,對於大範圍的註冊和驗證距離,放大失真對自動臉部識別系統效能的影響很小。
The magnification distortion is considered to be noticeable if the distance between units on a ruler at the nose tip level measured in pixels is more than 5%5 \% larger than the distance between units on a ruler at the outer canthus level measured in pixels. The elevation of the nose compared to the outer canthus of the test subject is assumed to be 50 mm . It is sufficient to measure this properly once whenever a photographic setup is introduced or modified. An example photo is given in Figure D.8. Examples of face portraits with good appearance and too strong magnification distortion are given in Figure D.7. The general case of the optical system is discussed in E.2. 如果以像素測量的鼻尖水平尺規單位間距離比以外眼角水平測量的尺規單位間距離大於 5%5 \% ,則認為放大失真明顯。測試對象的鼻子相對於外眼角的隆起假定為 50 毫米。只要引入或修改攝影設置,測量一次就足夠了。圖 D.8 中給出了範例照片。圖 D.7 中給出了外觀良好和放大失真過強的臉部肖像範例。光學系統的一般情況在 E.2 中討論。
Figure D. 8 - Magnification distortion measurement with rulers at nose and eye level 圖 D. 8 – 使用鼻部和眼部水平尺進行放大失真測量
There are several possible strategies for decreasing the magnification distortion. The general assessment of an optical system is discussed in E.2. Assuming a telecentric lens, the distance between sensor and subject does not introduce any magnification distortion. Real systems need specific considerations and measurements like those described in E.2. Another strategy to decrease the magnification distortion is to increase the distance between subject and camera, or to fold the optical path. The principle of a folded optical path is illustrated in Figure D.9. These strategies are not limitative. Camera subject distance and corresponding magnification distortion examples are listed in Table D.3. Sample images taken with a high quality camera with several magnification distortion rates are given in Figure D.7. 有幾種可能的策略可以減少放大失真。光學系統的總體評估在 E.2 中討論。假設使用遠心鏡頭,感測器與物體之間的距離不會引入任何放大失真。實際系統需要特定的考量和測量,例如 E.2 中所述的那些。另一種減少放大失真的策略是增加物體與相機之間的距離,或折疊光學路徑。折疊光學路徑的原理如圖 D.9 所示。這些策略並非限制性的。相機物體距離和相應的放大失真範例列於表 D.3 中。使用高品質相機拍攝的具有不同放大失真率的範例影像如圖 D.7 所示。
Figure D. 9 - Principle sketch of a folded optical path 圖 D. 9 - 折疊光路原理示意圖
Table D. 3 - Camera subject distance and corresponding magnification distortion 表 D. 3 – 相機物體距離和相應的放大失真
Camera subject distance in mm 相機主體距離,單位為毫米
Magnification distortion Deltad//d\Delta \mathrm{d} / \mathrm{d} for a standard (i.e. not telecentric) lens 標準(即非遠心)鏡頭的放大失真 Deltad//d\Delta \mathrm{d} / \mathrm{d}
300
16,7 %
400
12,5 %
500
10,0 %
600
8,3 %
700
7,1 %
1000
5,0 %
1200
4,2 %
1500
3,3 %
2000
2,5 %
2500
2,0 %
3000
1,7 %
Camera subject distance in mm Magnification distortion Deltad//d for a standard (i.e. not telecentric) lens
300 16,7 %
400 12,5 %
500 10,0 %
600 8,3 %
700 7,1 %
1000 5,0 %
1200 4,2 %
1500 3,3 %
2000 2,5 %
2500 2,0 %
3000 1,7 %| Camera subject distance in mm | Magnification distortion $\Delta \mathrm{d} / \mathrm{d}$ for a standard (i.e. not telecentric) lens |
| :--- | :--- |
| 300 | 16,7 % |
| 400 | 12,5 % |
| 500 | 10,0 % |
| 600 | 8,3 % |
| 700 | 7,1 % |
| 1000 | 5,0 % |
| 1200 | 4,2 % |
| 1500 | 3,3 % |
| 2000 | 2,5 % |
| 2500 | 2,0 % |
| 3000 | 1,7 % |
NOTE 2 This magnification distortion only applies for standard (i. e. non-telecentric lens) lenses. 註記 2 此放大失真僅適用於標準(即非遠心)鏡頭。
Homemade face portraits are also affected by magnification distortion tolerance. Issuers who accept homemade face portraits should be aware that there is no scientific solution that allows checking the tolerance compliancy. Therefore, the acceptance of homemade face portrait is not recommended. 自製臉部肖像也會受到放大失真容許度的影響。接受自製臉部肖像的發證者應注意,目前沒有科學方法可以檢查容許度是否符合規定。因此,不建議接受自製臉部肖像。
The issuer should allow for a transition period in which enrolment systems, e.g., photo booths and kiosks may be updated to fulfil the magnification distortion requirements considering economic and feasibility reasons. The duration of such transition period is at the discretion of the issuer. 發行者應考量經濟和可行性因素,允許一段過渡期,讓註冊系統,例如照相亭和自助服務機,能夠更新以符合放大失真要求。此過渡期的長度由發行者自行決定。
D.1.4.2.3 Radial distortion D.1.4.2.3 徑向失真
The radial distortion due to lens properties can be noticeable to human examiners but shall be within defined limits that allow effective face recognition. In particular, fish eye effects caused by wide angle lenses combined with camera placement too close to the face shall not be present. 由於鏡頭特性造成的徑向失真可能對人工檢查員而言顯而易見,但應在允許有效臉部辨識的定義限制內。特別是,廣角鏡頭與相機放置離臉部過近所造成的魚眼效果不應出現。
Acceptable distortion rate tolerances depend on the performance capacity of state of the art face recognition technology, and on the capability of typical human inspection staff to recognize people, even those coming from varying ethnic origin. 可接受的失真率容許度取決於最先進臉部辨識技術的效能,以及典型人工檢查人員辨識人員的能力,即使是來自不同種族的人員。
If the radial distortion is less than 2%2 \%, the human eye will not easily perceive it. It is recommended that radial distortion is less than 2,5%2,5 \%. 如果徑向畸變小於 2%2 \% ,人眼將不容易察覺。建議徑向畸變小於 2,5%2,5 \% 。
The general assessment of an optical system is discussed in E.2. E.2 討論了光學系統的一般評估。
D.1.4.2.4 Pixel count, focus and depth of field D.1.4.2.4 像素數、對焦與景深
Digital cameras used to capture face portraits shall produce images where the vertical and horizontal pixel density is the same. 用於拍攝臉部肖像的數位相機,其影像的垂直和水平像素密度應相同。
Live captured face portraits of a subject: 受測者即時擷取的人臉肖像:
Shall be captured in one of the following formats: PNG, JPEG, JPEG2000, RAW formats supported by the camera, lossless formats should be preferred, 應以以下其中一種格式擷取:PNG、JPEG、JPEG2000、相機支援的 RAW 格式,建議優先使用無損格式,
Should be captured at a minimum dimension of 1200 pixels width xx1600\times 1600 pixels height (cropped image), 應以最小尺寸擷取,寬度為 1200 像素 xx1600\times 1600 像素高度(裁切影像),
Shall be captured in colour. 應以彩色擷取。
One of the four possible encodings shall be used: 四種可能的編碼方式之一應被採用:
The JPEG sequential baseline (ISO/IEC 10918-1) mode of operation and encoded in the JFIF file format (the JPEG file format). JPEG 循序基準(ISO/IEC 10918-1)操作模式,並以 JFIF 檔案格式(JPEG 檔案格式)編碼。
The JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossy, and encoded in the JP2 file format (the JPEG2000 file format). JPEG-2000 第一部分碼流格式(ISO/IEC 15444-1),有損壓縮,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼。
The JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossless, and encoded in the JP2 file format (the JPEG2000 file format). JPEG-2000 第一部分碼流格式(ISO/IEC 15444-1),無損壓縮,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼。
The PNG specification (ISO/IEC 15948). PNG shall not be used in its interlaced mode and not for images that have been JPEG compressed before. PNG 規範 (ISO/IEC 15948)。PNG 不得使用其交錯模式,也不得用於先前已進行 JPEG 壓縮的影像。
For the use of RAW images see 7.40. The encoding into one of the four formats above can be done in a later process step before MRTD production. RAW 影像的使用請參閱 7.40。在 MRTD 製作之前,可於後續處理步驟中將影像編碼為上述四種格式之一。
The IED in the captured photo shall be at least 90 pixels for legacy applications. If an issuer considers the design of a new passport application process, the new IED should be at least 240 pixels. Examples for a new process could be live capturing, digital submission without analogue intermediate steps, or increasing the size of the printed photograph to be scanned, see Table D.4. See Figure 10 for an illustration of the IED measurement. 對於傳統應用,擷取照片中的 IED 應至少為 90 像素。如果發行者考慮設計新的護照申請流程,新的 IED 應至少為 240 像素。新流程的範例可能包括即時擷取、無類比中間步驟的數位提交,或增加要掃描的列印照片尺寸,請參閱表 D.4。IED 測量說明請參閱圖 10。
Criterion: live capture IED Requirement IED >= 90 pixels
Best practice IED >= 240 pixels
Criterion: scanned image IED Requirement IED >= 90 pixels
Best practice IED >= 240 pixels
Criterion: electronic submission IED Requirement IED >= 90 pixels
Best practice IED >= 240 pixels
Criterion: issuer repository IED Requirement IED >= 90 pixels
Best practice IED >= 240 pixels
Criterion: MRTD chip storage IED Requirement IED >= 90 pixels
Best practice IED >= 120 pixels| Criterion: live capture IED | Requirement | IED $\geq 90$ pixels |
| :--- | :--- | :--- |
| | Best practice | IED $\geq 240$ pixels |
| Criterion: scanned image IED | Requirement | IED $\geq 90$ pixels |
| | Best practice | IED $\geq 240$ pixels |
| Criterion: electronic submission IED | Requirement | IED $\geq 90$ pixels |
| | Best practice | IED $\geq 240$ pixels |
| Criterion: issuer repository IED | Requirement | IED $\geq 90$ pixels |
| | Best practice | IED $\geq 240$ pixels |
| Criterion: MRTD chip storage IED | Requirement | IED $\geq 90$ pixels |
| | Best practice | IED $\geq 120$ pixels |
NOTE This pixel count is specified for the live captured face portrait only. For stored images on the chip, see D.1.5 and Table D. 10 as well. 附註:此像素計數僅針對即時擷取的人臉肖像指定。至於晶片上儲存的影像,另請參閱 D.1.5 和表 D.10。
All images shall have sufficient focus and depth of field to maintain the required level of details. The camera shall be capable of accurately rendering fine contrasted face details, such as wrinkles and moles, as small as 1 mm in diameter on the face. 所有影像應具備足夠的焦距和景深,以維持所需的細節程度。相機應能準確呈現臉部細微對比的細節,例如臉上直徑小至 1 毫米的皺紋和痣。
The focus and depth of field of the camera shall be set so that the subject’s captured image is in focus from nose to ears. In most cases, a depth of field of 150 mm will be sufficient. See Table D.5. The background behind the subject may be out of focus. Proper focus and depth-of-field will be assured by either using the camera auto focus function with manual aperture settings or by pre-focusing the lens at the distance of the subject’s eyes and by selecting an appropriate aperture ( F -stop) to ensure a depth-of-field of the distance from a subject’s nose to ears. See E.5. 相機的對焦和景深應設定為使拍攝對象的影像從鼻子到耳朵都清晰。在大多數情況下,150 毫米的景深就足夠了。請參閱表 D.5。拍攝對象後方的背景可以失焦。透過使用手動光圈設定的相機自動對焦功能,或將鏡頭預先對焦在拍攝對象眼睛的距離,並選擇適當的光圈(F 值)以確保從拍攝對象鼻子到耳朵的景深,將可確保適當的對焦和景深。請參閱 E.5。
Table D. 5 - Depth of field requirements and recommendations 表 D.5 — 景深要求與建議
準則:景深
Criterion:
Depth of field
Criterion:
Depth of field| Criterion: |
| :--- |
| Depth of field |
Requirement 要求
Nose to ears 從鼻尖到耳際
Best practice 最佳實務
150 mm from nose level 距離鼻尖 150 毫米
"Criterion:
Depth of field" Requirement Nose to ears
Best practice 150 mm from nose level| Criterion: <br> Depth of field | Requirement | Nose to ears |
| :--- | :--- | :--- |
| | Best practice | 150 mm from nose level |
EXAMPLE An aperture of f//8\mathrm{f} / 8 for an 80 mm lens at a distance of 2500 mm provides a depth of field of 150 mm . An aperture of f//16\mathrm{f} / 16 for a 50 mm lens at a distance of 1200 mm provides a depth of field of about 180 mm . 範例 在 2500 毫米的距離下,80 毫米鏡頭的 f//8\mathrm{f} / 8 光圈可提供 150 毫米的景深。在 1200 毫米的距離下,50 毫米鏡頭的 f//16\mathrm{f} / 16 光圈可提供約 180 毫米的景深。
A simplified visual compliance check method requires that the individual millimetre markings of rulers placed on the subject’s nose and ear facing the camera can be seen simultaneously in a captured test image. See Figure D.10. This quality assurance method should be used for quality assurance field checks from time to time. A more systematic test method is described in E.2. 一種簡化的視覺合規性檢查方法要求,在拍攝的測試影像中,可以同時看到放置在受試者鼻子和耳朵上、面向攝影機的尺子上的個別毫米標記。請參閱圖 D.10。這種品質保證方法應不時用於品質保證現場檢查。更系統的測試方法請參閱 E.2。
Figure D. 10 - Examples for sharpness at nose and ear level 圖 D.10 - 鼻尖和耳部清晰度的範例
D.1.4.2.5 Background D.1.4.2.5 背景
The background surface behind the subject shall be plain, and shall have no texture containing spots, lines or curves that will be visible in the captured image. The background shall have a uniform colour. There may be gradual changes from light to dark luminosity in a single direction, although this may make it more difficult to remove the background during the document production process. 主體後方的背景表面應為素色,且不得有在拍攝影像中可見的斑點、線條或曲線等紋理。背景應為單一均勻的顏色。背景亮度可沿單一方向由亮漸變至暗,儘管這可能會增加文件製作過程中移除背景的難度。
A typical background for the scene is grey with a plain, dull flat surface. Plain light-coloured backgrounds such as light blue or white may be used as long as there is sufficient distinction between the face/hair area and the background. Camera colour settings should not be shifted depending on the background colour, see Figure D.11. 場景的典型背景為灰色,表面平坦、無光澤。只要臉部/頭髮區域與背景之間有足夠的區別,亦可使用淺色背景,例如淺藍色或白色。相機色彩設定不應隨背景顏色而改變,請參閱圖 D.11。
Figure D. 11 - Examples of compliant portrait backgrounds 圖 D. 11 - 符合規範的人像背景範例
The boundary between the head and the background should be clearly identifiable around the entire subject with the exception of very large hair volume. See Figure D.12. A boundary that is not clearly visible can have a negative impact on the production process which often requires background removal. 頭部與背景之間的邊界應在整個主體周圍清晰可辨,但髮量非常大的情況除外。請參閱圖 D.12。邊界不清晰可能會對生產過程產生負面影響,因為生產過程通常需要移除背景。
Figure D. 12 - Contrast examples 圖 D.12 - 對比範例
Shadows should not be visible on the background behind the face image. In particular, there shall not be asymmetric shadows. There shall not be any objects visible in the background like supporting persons, chair backs, furniture, carpets, patterned wall papers or plants. For examples, see Figure D.13. 臉部影像後方的背景上不應出現陰影。特別是,不應出現不對稱的陰影。背景中不應出現任何可見的物體,例如支撐人員、椅背、家具、地毯、有圖案的壁紙或植物。範例請參閱圖 D.13。
Figure D. 13 - Examples for non-compliant backgrounds 圖 D.13 - 不符合規定的背景範例
D.1.4.2.6 Lighting D.1.4.2.6 照明
Face portraits shall have adequate and uniform illumination. Lighting shall be equally distributed on the face, in particular symmetrically, i.e., there is no difference between the brightness of the right and left side of the face. There shall not be significant direction of the light visible from the point of view of the camera. 臉部肖像應有充足且均勻的光線。光線應均勻分佈於臉部,特別是對稱分佈,即臉部左右兩側的亮度不應有差異。從攝影機的角度來看,不應有明顯的光線方向。
The measured EV at four spots on a subject’s face; the left and right cheeks, forehead, and chin, should be the same. An EV difference of at most one F-stop or one shutter speed step is acceptable. If one or some of these four spots are covered by hair, e.g., the forehead by the hairstyle or the chin by a beard, these spots can’t be evaluated. The appropriate illumination setup of the scene should be verified from time to time. The subject being used for these tests should not have a hairstyle covering the forehead or the cheeks, or a beard. 受測者臉部四個點(左右臉頰、額頭和下巴)的測量曝光值(EV)應相同。曝光值差異最多一個光圈級數或一個快門速度級數是可以接受的。如果這四個點中的一個或幾個被頭髮遮蓋,例如額頭被髮型遮蓋或下巴被鬍鬚遮蓋,則這些點無法進行評估。應不時驗證場景的適當照明設置。用於這些測試的受測者不應有遮蓋額頭或臉頰的髮型,或留有鬍鬚。
The uniformity measurement should be done as specified below. It is not intended to be used for every single image. See Figure D. 14 for a visualization of that measurement. Automated quality assurance software, e.g., for registration offices or photo kiosks, should be implemented accordingly. However, such software should also consider exceptions due to hair on the forehead, beard, face anomalies and the like. 均勻度測量應按以下規定進行。它不適用於每一張單獨的影像。請參閱圖 D.14 以視覺化該測量。自動品質保證軟體,例如用於登記處或照相亭的軟體,應據此實施。然而,此類軟體也應考慮因額頭上的頭髮、鬍鬚、臉部異常等造成的例外情況。
Determine the four measurement zones on the forehead, the cheeks and the chin. These locations are determined as follows: 確定額頭、臉頰和下巴的四個測量區域。這些位置的確定方式如下:
a) Connect the two eye centres (feature points 12.1 and 12.2 from ISO/IEC 14496-2:2004, Annex C). The IED is the length of the connecting line HH. The point MM is the midpoint of this line. a) 連接兩個眼球中心(ISO/IEC 14496-2:2004,附錄 C 中的特徵點 12.1 和 12.2)。IED 是連接線 HH 的長度。點 MM 是這條線的中點。
b) Connect M with the mouth midpoint (feature point 2.3 from ISO/IEC 14496-2:2004, Annex C). EMD is the length of the connecting line V. Note that the two lines do not need to be rectangular. b) 將 M 與嘴巴中點連接(ISO/IEC 14496-2:2004,附錄 C 中的特徵點 2.3)。EMD 是連接線 V 的長度。請注意,這兩條線不需呈直角。
c) MP , the side length of the four squared measurement zones, is defined to be 0,3 IED. c) MP,即四個方形測量區域的邊長,定義為 0.3 IED。
d) The centre of the forehead measurement zone F is located at a distance of 0,5 EMD upwards from M on V . d) 額頭測量區 F 的中心位於 M 點上方 0.5 EMD 距離處,在 V 軸上。
e) The centre of the chin measurement zone C is located at a distance of 1,5 EMD downwards from M on V . e) 下巴測量區 C 的中心位於 M 點下方 1.5 EMD 距離處,在 V 軸上。
f) The top left corner of the right (from the capture subject) cheek measurement zone R is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the left of M on H (looking at the capture subject). f) 右臉頰(從拍攝對象角度看)測量區 R 的左上角位於 M 點下方 0.5 EMD 距離處,在 V 軸上,以及 M 點左側 0.5 IED 距離處,在 H 軸上(從拍攝對象角度看)。
g) The top right corner of the left (from the capture subject) cheek measurement zone LL is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the right of M on H (looking at the capture subject). g) 左臉頰(從拍攝對象角度看)測量區 LL 的右上角位於 M 點下方 0.5 EMD 距離處,在 V 軸上,以及 M 點右側 0.5 IED 距離處,在 H 軸上(從拍攝對象角度看)。
For all colour channels, measure the mean intensity values MI for the measurement zones F,C\mathrm{F}, \mathrm{C}, L and R . 針對所有色彩通道,測量測量區域 F,C\mathrm{F}, \mathrm{C} 、L 和 R 的平均強度值 MI。
For all channels separately, the lowest MI (of F, C, L and R) in that channel shall not be lower than 50 % of the highest MI (of F, C, L and R). 針對所有通道,個別而言,該通道中最低的 MI(F、C、L 和 R 中)不得低於最高 MI(F、C、L 和 R 中)的 50%。
Figure D. 14 - Location and size of the intensity measurement zones 圖 D.14 - 強度測量區域的位置和大小
The measures for the illumination intensity and requirements to them are listed in Table D.6. 照明強度的測量方法及其要求列於表 D.6。
Table D. 6 - Measures for the illumination intensity compliance check 表 D. 6 — 照明強度合規性檢查的量測
Term 術語
Description 說明
Requirement 要求
12.1
Feature point at left eye centre 左眼中心特徵點
12.2
Feature point at right eye centre 右眼中心特徵點
H
Line connecting 12.1 and 12.2 連接 12.1 和 12.2 的線
IED
Length of H between 12.1 and 12.2 12.1 和 12.2 之間 H 的長度
IED >= 90\geq 90 pixels IED >= 90\geq 90 像素
M
Midpoint of H between 12.1 and 12.2 H 點介於 12.1 和 12.2 之間的中點
2.3
Feature point at mouth centre (with closed mouth the same as 2.2) 嘴巴中心特徵點(閉嘴時與 2.2 相同)
V
Line connecting M and 2.3, V and H do not need to be orthogonal 連接 M 和 2.3 的線,V 和 H 不需正交
EMD
Length of V between M and 2.3 M 和 2.3 之間 V 的長度
MP
Side length of the squared measurement zones 方形測量區域的邊長
MP=0,3IED\mathrm{MP}=0,3 \mathrm{IED}
F
Forehead measurement zone, located at a distance of 0,5 EMD upwards from M on V 額頭測量區域,位於 V 軸上 M 點上方 0.5 EMD 的距離處
C
Chin measurement zone C , located at a distance of 1,5 EMD downwards from M on V 下巴測量區域 C,位於 V 軸上 M 點下方 1.5 EMD 的距離處
R
Right (from the capture subject) cheek measurement zone R, its top left corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the left of M on H 右(從拍攝對象角度)臉頰測量區域 R,其左上角位於 V 軸上 M 點下方 0.5 EMD 的距離處,以及 H 軸上 M 點左側 0.5 IED 的距離處
Term Description Requirement
12.1 Feature point at left eye centre
12.2 Feature point at right eye centre
H Line connecting 12.1 and 12.2
IED Length of H between 12.1 and 12.2 IED >= 90 pixels
M Midpoint of H between 12.1 and 12.2
2.3 Feature point at mouth centre (with closed mouth the same as 2.2)
V Line connecting M and 2.3, V and H do not need to be orthogonal
EMD Length of V between M and 2.3
MP Side length of the squared measurement zones MP=0,3IED
F Forehead measurement zone, located at a distance of 0,5 EMD upwards from M on V
C Chin measurement zone C , located at a distance of 1,5 EMD downwards from M on V
R Right (from the capture subject) cheek measurement zone R, its top left corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the left of M on H | Term | Description | Requirement |
| :--- | :--- | :--- |
| 12.1 | Feature point at left eye centre | |
| 12.2 | Feature point at right eye centre | |
| H | Line connecting 12.1 and 12.2 | |
| IED | Length of H between 12.1 and 12.2 | IED $\geq 90$ pixels |
| M | Midpoint of H between 12.1 and 12.2 | |
| 2.3 | Feature point at mouth centre (with closed mouth the same as 2.2) | |
| V | Line connecting M and 2.3, V and H do not need to be orthogonal | |
| EMD | Length of V between M and 2.3 | |
| MP | Side length of the squared measurement zones | $\mathrm{MP}=0,3 \mathrm{IED}$ |
| F | Forehead measurement zone, located at a distance of 0,5 EMD upwards from M on V | |
| C | Chin measurement zone C , located at a distance of 1,5 EMD downwards from M on V | |
| R | Right (from the capture subject) cheek measurement zone R, its top left corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the left of M on H | |
Table D. 6 (continued) 表 D. 6 (續)
Term 術語
Description 說明
Requirement 要求
L
Left (from the capture subject) cheek measurement zone L, its top right corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the right of M on H 左(從擷取對象看)臉頰測量區 L,其右上角位於從 M 點沿 V 軸向下 0.5 EMD 處,以及從 M 點沿 H 軸向右 0.5 IED 處
MI
Mean intensity value measured for every channel separately 每個通道分別測量的平均強度值
Term Description Requirement
L Left (from the capture subject) cheek measurement zone L, its top right corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the right of M on H
MI Mean intensity value measured for every channel separately max <= 2xx min (per channel)| Term | Description | Requirement |
| :--- | :--- | :--- |
| L | Left (from the capture subject) cheek measurement zone L, its top right corner is located at a distance of 0,5 EMD downwards from M on V and 0,5 IED to the right of M on H | |
| MI | Mean intensity value measured for every channel separately | max $\leq 2 \times \min$ (per channel) |
While it is understood that massive shadows on parts of the face will obscure face details important for identification, having no shadows at all will result in a non-natural appearance. In such a case, the face will appear flat and without surface features. Appropriate shadows help distinguish the shape of the nose, eye areas, forehead, cheeks, chin and so on. Furthermore, lighting and shadows are necessary to show details around the eyes, wrinkles and scars. There shall not be extreme dark shadow visible on the face, especially around the nose, in the eye sockets, around the mouth, and between mouth and chin that obscure face details important for inspection. The brightness shall be nearly the same on both sides of the face, left and right. All features in the face shall be clearly recognizable, and the volume effect especially around nose and eyes shall render the reality, see Figure D.15. 雖然可以理解臉部某些部位的大片陰影會遮蔽對辨識很重要的臉部細節,但完全沒有陰影會導致不自然的樣貌。在這種情況下,臉部會顯得扁平且沒有表面特徵。適當的陰影有助於區分鼻子、眼睛區域、額頭、臉頰、下巴等的形狀。此外,光線和陰影對於顯示眼睛周圍的細節、皺紋和疤痕是必要的。臉部不應出現極端黑暗的陰影,特別是鼻子周圍、眼窩、嘴巴周圍以及嘴巴和下巴之間,這些陰影會遮蔽對檢查很重要的臉部細節。臉部左右兩側的亮度應幾乎相同。臉部所有特徵應清晰可辨,且體積效果,特別是鼻子和眼睛周圍,應能呈現真實感,請參閱圖 D.15。
EXAMPLE To comply with this requirement, the illumination elements can be aligned in an angle of approximately 35^(@)35^{\circ} off the axis between camera lens and face centre. Descriptions of sample illumination layouts are given in E.1. 範例 為符合此要求,照明元件可沿攝影機鏡頭與臉部中心之間軸線約 35^(@)35^{\circ} 的角度對齊。範例照明佈局的說明請參閱 E.1。
a) Side illumination a) 側邊照明
b) Top illumination b) 頂部照明
c) Bottom illumination c) 底部照明
Figure D. 15 - Examples for non-compliant illumination 圖 D.15 — 不符合規範照明的範例
Flashes should only be used for indirect illumination. Issuers may exclude the use of flashes. If face portraits are captured using flashes, care should be taken to verify that the eyes of the subject are open. As long as the requirements for the face portrait from D.1.4 are maintained, one or more flashes or a large surface flash may be used. There shall not be any shadows at the face or in the background of the face portrait that obscure face details important for inspection. Illumination shall not cause any red eye effect visible in the eyes or other lighting artefacts such as spots from a ring flash reducing the visibility of the eyes. 閃光燈僅應用於間接照明。發證機關可排除使用閃光燈。如果臉部肖像使用閃光燈拍攝,應注意確認受試者的眼睛是睜開的。只要符合 D.1.4 中臉部肖像的要求,可以使用一個或多個閃光燈或大型表面閃光燈。臉部或臉部肖像背景不應有任何陰影遮蔽對檢查重要的臉部細節。照明不應導致眼睛出現任何紅眼效應或其他照明偽影,例如環形閃光燈造成的斑點,從而降低眼睛的可見度。
A high colour rendering index is recommended for illumination. See D.1.4.2.9 for details. 建議照明採用高顯色指數。詳情請參閱 D.1.4.2.9。
The captured image shall contain minimal reflections or bright spots. Diffused lighting, multiple balanced sources or other appropriate lighting methods should be used. A single bare point light source like a camera mounted direct flash shall not be used for imaging. Lamp reflectors or other technologies that provide non-point illumination should be used. 拍攝的影像應盡量減少反光或亮點。應使用漫射照明、多個平衡光源或其他適當的照明方法。不應使用單一裸露的點光源(例如相機上安裝的直射閃光燈)進行成像。應使用燈具反射器或其他提供非點狀照明的技術。
D.1.4.2.7 Contrast D.1.4.2.7 對比度
For each patch of skin on the capture subject’s face, the gradations in textures shall be clearly visible, i.e., being of reasonable contrast. Whites of eyes shall be clearly light or white (when appropriate) and 對於拍攝對象臉部上的每個皮膚區域,紋理的漸變應清晰可見,即具有合理的對比度。眼睛的白色部分應清晰地呈現淺色或白色(適當時),且
dark hair or face features (when appropriate) shall be clearly dark. Generally, the face portrait shall have appropriate brightness and good contrast between face, hair and background. See Figure D.16. 深色頭髮或臉部特徵(適當時)應清晰地呈現深色。一般而言,臉部肖像應具有適當的亮度,以及臉部、頭髮和背景之間良好的對比度。請參閱圖 D.16。
Figure D. 16 - Examples for compliant and non-compliant exposure 圖 D. 16 - 符合與不符合曝光的範例
D.1.4.2.8 Dynamic range D.1.4.2.8 動態範圍
The dynamic range of the image should have at least 50%50 \% of intensity variation in the face region of the image. The face region is defined as the region from crown to chin and from the left ear to the right ear. This recommendation may require an adjustment of the equipment settings on an individual basis when the skin tone is excessively light or dark. In the rectangle between the ISO/IEC 14496-2 feature points: 影像的動態範圍在臉部區域應至少有 50%50 \% 的強度變化。臉部區域定義為從頭頂到下巴,以及從左耳到右耳的區域。當膚色過於淺或深時,此建議可能需要根據個人情況調整設備設定。在 ISO/IEC 14496-2 特徵點之間的矩形內:
2.1: Bottom of the chin, 2.1:下巴底部,
10.9: Upper contact point between left ear and face, 10.9:左耳與臉部之間的上方接觸點,
10.10: Upper contact point between right ear and face, and 10.10:右耳與臉部之間的上方接觸點,以及
11.1: Middle border between hair and forehead. 11.1:頭髮與額頭之間的中間邊界。
All colour channels should have at least 50%50 \% of intensity variation. As this may be difficult to achieve, best efforts should be made to get as close as possible to that requirement. See Figure D. 17 for an illustration of the recommended measuring zone and Figure D. 18 for examples of good and bad quality images. 所有色彩通道的強度變化應至少為 50%50 \% 。由於這可能難以實現,應盡力接近該要求。請參閱圖 D.17 以了解建議的測量區域,並參閱圖 D.18 以了解良好和不良品質影像的範例。
Figure D. 17 - Recommended dynamic range measuring zone 圖 D. 17 - 建議的動態範圍測量區域
a) Compliant face portrait a) 符合規範的臉部肖像
b) Too low dynamic range b) 動態範圍過低
Figure D. 18 - Compliant and non-compliant dynamic range 圖 D. 18 - 符合與不符合規範的動態範圍
D.1.4.2.9 Colour D.1.4.2.9 顏色
All images should be captured in colour. Newly designed enrolment processes should capture colour images only. 所有影像應以彩色擷取。新設計的註冊流程應僅擷取彩色影像。
The captured face portrait shall be a true-colour representation of the holder in a typical colour space such as sRGB as specified in IEC 61966-2. Other true-colour representations may be used as long as the colour profile is embedded in the image. 擷取的人臉肖像應為持有者的真彩色表示,採用典型色彩空間,例如 IEC 61966-2 中指定的 sRGB。只要色彩設定檔嵌入影像中,即可使用其他真彩色表示。
The sensor of the camera shall capture the entire visible wavelength, basically the wavelength between 400 nm and 700 nm . It allows rendering correctly the natural colours seen by humans. Unnaturally coloured lighting, i.e., yellow, red, etc., shall not be used. Care should be taken to correct the white balance of image capture devices. The lighting shall produce a face image with naturally looking flesh tones when viewed in typical examination environments. See Figure D.19. 相機的感測器應擷取整個可見光波長,基本上是 400 奈米到 700 奈米之間的波長。這使得能夠正確呈現人類所見的自然色彩。不應使用非自然色彩的照明,例如黃色、紅色等。應注意校正影像擷取裝置的白平衡。照明應在典型的檢查環境中觀看時,產生具有自然膚色的人臉影像。請參閱圖 D.19。
Figure D. 19 - Examples for compliant and non-compliant colour setups 圖 D. 19 - 符合與不符合規範的色彩設定範例
The RGB values from the capturing device should be converted to an appropriate RGB space as required by the data format. 擷取裝置的 RGB 值應依資料格式要求轉換為適當的 RGB 空間。
Dedicated near infra-red cameras shall not be used for image acquisition. 專用近紅外線攝影機不得用於影像擷取。
Colour calibration using an 18 % grey background or other method such as white balancing should be applied. 應使用 18% 灰度背景或其他方法(例如白平衡)進行色彩校準。
White balance shall be properly set in order to achieve high fidelity skin tones. Quality assurance measurements of light conditions and camera system response should be made when a recommended CIE Standard Illuminant D65 illuminant (see ISO 11664-2) or a similar continuous spectrum daylight illuminant and a camera and/or camera control software are used to take pictures. In practice it is necessary to reduce the ambient light emanating from uncontrolled daylight sources, fluorescent or similar light sources and reflections from surfaces. 應正確設定白平衡,以實現高傳真膚色。當使用建議的 CIE 標準光源 D65(參見 ISO 11664-2)或類似的連續光譜日光光源,以及相機和/或相機控制軟體進行拍攝時,應對光線條件和相機系統回應進行品質保證測量。實際上,有必要減少來自不受控制的日光光源、螢光燈或類似光源以及表面反射的環境光。
Imaging fidelity measurements for photo studio and stationary registration office installations may be done either using a light spectrum analyser to define the spectral characteristics of the illuminants or analysing measurement target images using software applications. 攝影棚和固定式登記處裝置的影像傳真度測量,可透過使用光譜分析儀來定義光源的光譜特性,或使用軟體應用程式分析測量目標影像來完成。
Annex E. 3 contains a methodology for measuring colour quality and recommended values. 附錄 E.3 包含測量色彩品質的方法和建議值。
Colour quality should be measured in terms of colour error using the CIEDE2000 formula (deltaE2000) of a standardized test pattern according to the methodology in E.3. The average deltaE2000 of all colour patches should not exceed 4 for scanners and 10 for camera systems. The maximum deltaE2000 for any colour patch should not exceed 15 for scanners and 20 for camera systems. Measured CIELAB Lab* human skin tone a^(**)a^{*} and b^(**)b^{*} values shall be positive as shown in E.3. See References [25] and [26] for explanations. Negative a^(**)a^{*} and b^(**)b^{*} values are acceptable for medical reasons only. 色彩品質應根據 E.3 中的方法,使用標準化測試圖案的 CIEDE2000 公式(deltaE2000)以色彩誤差來測量。所有色塊的平均 deltaE2000,掃描器不應超過 4,相機系統不應超過 10。任何色塊的最大 deltaE2000,掃描器不應超過 15,相機系統不應超過 20。測量的 CIELAB Lab*人類膚色 a^(**)a^{*} 和 b^(**)b^{*} 值應為正值,如 E.3 所示。有關說明請參見參考文獻[25]和[26]。負值 a^(**)a^{*} 和 b^(**)b^{*} 僅在醫療原因下可接受。
D.1.4.2.10 Noise D.1.4.2.10 雜訊
The enrolment should be made in a controlled scene; the picture should be captured with high signal-to-noise ratio. Noise is not information contained in the original scene but is created by the electronics due to a too high level of amplification. ISO sensitivity settings at values of ISO 100 and ISO 200 typically 註冊應在受控場景中進行;圖片應以高訊號雜訊比擷取。雜訊並非原始場景中包含的資訊,而是由於過高的放大程度而由電子設備產生。ISO 感光度設定在 ISO 100 和 ISO 200 的值通常
reduce noise; for high-quality cameras ISO 400 and ISO 800 may also be used. Noise can be minimized by correct exposure at a low ISO setting. 可減少雜訊;對於高品質相機,ISO 400 和 ISO 800 也可使用。透過在低 ISO 設定下正確曝光,可將雜訊降至最低。
The ratio of signal to noise (SNR) is one indicator of the overall ability of a collection system to accurately capture a subject’s appearance. Unwanted variations in the response of an imaging system (i. e., noise) are inherent in the capture process of a digital representation of a physical scene and arise from the interplay between the system components (e.g., sensor and lens) and the capture environment (e. g., subject illumination). Reducing overall noise to improve the SNR benefits human examiners and automated face analysis systems which rely on high-quality subject images. SNR should be computed as prescribed in ISO 15739:2017, 4.7, which incorporates a human visual model to calculate the human observable (i. e., perceived) SNR of the overall collection system. 訊號雜訊比 (SNR) 是衡量採集系統準確捕捉主體外觀整體能力的指標之一。影像系統回應中不必要的變異(即雜訊)是物理場景數位表示擷取過程中固有的,並且源於系統組件(例如,感測器和鏡頭)與擷取環境(例如,主體照明)之間的相互作用。降低整體雜訊以提高 SNR 有益於人類檢查員和自動化臉部分析系統,這些系統依賴高品質的主體影像。SNR 應按照 ISO 15739:2017, 4.7 的規定計算,該標準結合了人類視覺模型來計算整體採集系統的人類可觀察(即感知)SNR。
Commercial software designed for use by photo studios and registration office imaging systems are available with accompanying standard test targets for computing SNR. 供照相館和登記處影像系統使用的商業軟體,可搭配標準測試目標來計算訊噪比。
D.1.4.2.11 Filters D.1.4.2.11 濾鏡
Polarization filters shall not be used in front of the light sources. Linear polarization filters shall not be used in front of the camera lens as they interfere with autofocus cameras and thus reduce or remove skin texture information which might be used by face image comparison algorithms. Circular polarizing filters decrease reflections that show up in eyeglasses and may be used in front of the camera lens. 光源前方不得使用偏光濾鏡。相機鏡頭前方不得使用線性偏光濾鏡,因為它們會干擾自動對焦相機,進而減少或消除臉部影像比對演算法可能使用的皮膚紋理資訊。圓形偏光濾鏡可減少眼鏡上出現的反光,並可用於相機鏡頭前方。
D.1.4.3 Subject conditions D.1.4.3 主體條件
D.1.4.3.1 Pose D.1.4.3.1 姿勢
The subject should be instructed to look directly at the camera and to keep his or her head erect. Typically people are able to adopt such a position if instructed. Care should be taken to maintain the full frontal pose as well as possible. See Figure D.20. 應指示受測者直視攝影機並保持頭部挺直。通常,如果受指示,人們能夠採取這種姿勢。應注意盡可能保持完全正面的姿勢。請參閱圖 D.20。
Figure D. 20 - Pose examples 圖 D.20 - 姿勢範例
The shoulders shall be square on to the camera, parallel to the camera imaging plane. Portrait style photographs where the subject is looking over the shoulder shall not be used. See Figure D.21. 肩膀應與攝影機呈直角,平行於攝影機成像平面。不應使用受測者回頭看的肖像風格照片。請參閱圖 D.21。
Figure D. 21 - Pose examples 圖 D. 21 - 姿勢範例
The pitch of the head shall be less than +-5^(@)\pm 5^{\circ} from frontal. The yaw of the head shall be less than +-5^(@)\pm 5^{\circ} from frontal. The roll of the head shall be less than +-8^(@)\pm 8^{\circ}, it is recommended to keep it below +-5^(@)\pm 5^{\circ}. Any stronger pose deviation may have negative impact on face recognition error rates. Therefore, effort should be spent to ensure that all angles are as small as possible. See Table D.7. For an illustration of the angles see Figure 3. For samples showing correct pose and pose deviations see Figure D.22. 頭部的俯仰角應小於正面 +-5^(@)\pm 5^{\circ} 。頭部的偏航角應小於正面 +-5^(@)\pm 5^{\circ} 。頭部的滾轉角應小於 +-8^(@)\pm 8^{\circ} ,建議保持在 +-5^(@)\pm 5^{\circ} 以下。任何更強烈的姿勢偏差都可能對臉部辨識錯誤率產生負面影響。因此,應努力確保所有角度都盡可能小。請參閱表 D.7。角度的圖示請參閱圖 3。顯示正確姿勢和姿勢偏差的範例請參閱圖 D.22。
Table D. 7 - Pose angle requirements and recommendations 表 D. 7 - 姿勢角度要求與建議
Criterion: Requirement Pitch <= +-5^(@), yaw <= +-5^(@), roll <= +-8^(@)
Pose angle Best Practice Pitch <= +-5^(@), yaw <= +-5^(@), roll <= +-5^(@)| Criterion: | Requirement | Pitch $\leq \pm 5^{\circ}$, yaw $\leq \pm 5^{\circ}$, roll $\leq \pm 8^{\circ}$ |
| :--- | :--- | :--- |
| Pose angle | Best Practice | Pitch $\leq \pm 5^{\circ}$, yaw $\leq \pm 5^{\circ}$, roll $\leq \pm 5^{\circ}$ |
Figure D. 22 - Pose angle examples 圖 D. 22 - 姿態角範例
D.1.4.3.2 Expression D.1.4.3.2 表情
The face shall have a neutral expression; in particular the capture subject shall not smile. The mouth shall be closed; the teeth shall not be visible. A smile is not allowed, even with closed jaw. The eyebrows shall not be raised. Squinting and frowning shall not be visible. See Figure D. 23. 臉部應保持中性表情;特別是拍攝對象不應微笑。嘴巴應閉合;牙齒不應露出。不允許微笑,即使是閉合下顎。眉毛不應上揚。不應出現瞇眼和皺眉的表情。請參閱圖 D. 23。
Figure D. 23 - Expression examples 圖 D. 23 - 表情範例
The mouth is considered to be closed if the distance A between the inner borders of the lips (distance between feature points 2.2 and 2.3) is less than 50 % of the thickness of the lower lip B (distance between the feature points 2.3 and 8.2). See Figure D.24. 如果嘴唇內緣之間的距離 A(特徵點 2.2 和 2.3 之間的距離)小於下唇厚度 B(特徵點 2.3 和 8.2 之間的距離)的 50%,則視為閉合的嘴巴。請參閱圖 D.24。
Figure D. 24 - Definition of a closed mouth 圖 D. 24 - 閉合嘴巴的定義
D.1.4.3.3 Eye visibility D.1.4.3.3 眼睛可見度
Both eyes shall be opened naturally, but not forced wide-opened. Pupils and irises, including iris colour, shall be completely visible, although there may be exceptions due to ethnicity or other individually specific reasons. The eyes shall look into the camera unless there are medical conditions preventing this. There should not be strong shadows in the eye-sockets. See Figure D. 25 for examples. 雙眼應自然睜開,但不可強行撐大。瞳孔和虹膜,包括虹膜顏色,應完全可見,儘管可能因種族或其他個人特定原因而有例外。眼睛應直視攝影機,除非有醫療狀況阻礙。眼窩不應有強烈陰影。範例請參閱圖 D. 25。
Figure D. 25 - Eye visibility examples 圖 D. 25 - 眼睛可見度範例
Any lighting artefacts present on the region of the eyes shall not obscure eye details such that identification becomes difficult. Lighting artefacts shall not be larger than 15%15 \% of the area of the iris. If there are unacceptable reflections, the illumination should be relocated appropriately. The pitch shall not be increased by moving the head forward. 眼睛區域上出現的任何光學偽影,不應模糊眼睛細節,以免辨識困難。光學偽影不應大於虹膜面積的 15%15 \% 。如果出現無法接受的反光,應適當調整照明位置。不應透過將頭部前傾來增加音高。
Examples of setups preventing or at least reducing lighting artefacts are given in E.1. 防止或至少減少照明偽影的設置範例請參閱 E.1。
The eye visibility zone (EVZ) is defined as the covering rectangle having a distance V of at least 5%5 \% of the IED to any part of the visible eye ball. Figure D. 26 indicates the distance V of the covering rectangle to the visible parts of the eye ball. The EVZ shall be completely visible and unobscured. 眼睛可見區域 (EVZ) 定義為覆蓋矩形,其與 IED 的距離 V 至少為 5%5 \% ,且與可見眼球的任何部分保持距離。圖 D.26 顯示了覆蓋矩形與眼球可見部分的距離 V。EVZ 應完全可見且無遮蔽。
Figure D. 26 - Illustration of the EVZ 圖 D. 26 - EVZ 示意圖
Contact lenses changing the appearance of the iris including the size and the shape shall not be worn. The pattern of the lens shall not exceed the limbus. 不得配戴會改變虹膜外觀(包括大小和形狀)的隱形眼鏡。鏡片的圖案不得超出角膜緣。
D.1.4.3.4 Accessories: Glasses D.1.4.3.4 配件:眼鏡
If glasses are permitted by the issuer, subjects may wear glasses during image capture if they typically do so. Glasses other than those worn due to ametropia shall not be worn. Reading glasses shall not be worn during image capture. The lens area of glasses shall be made of fully transparent material. Tinted glasses, sunglasses, and glasses with polarization filters shall not be worn. An exception applies when the subject asserts a medical reason to retain glasses which are not fully transparent. If glasses are worn that tint automatically under illumination, they shall be photographed without tint by tuning the direct illumination or background lighting. In cases where the tint cannot be reduced the glasses shall 如果發證機關允許,受測者在影像擷取時可配戴眼鏡,如果他們通常會配戴。除了因屈光不正而配戴的眼鏡外,不得配戴其他眼鏡。在影像擷取時不得配戴老花眼鏡。眼鏡的鏡片區域應由完全透明的材料製成。不得配戴有色眼鏡、太陽眼鏡和帶偏光濾鏡的眼鏡。當受測者聲稱有醫療原因必須保留非完全透明的眼鏡時,則適用例外情況。如果配戴的眼鏡在光照下會自動變色,則應透過調整直接照明或背景照明來拍攝無色的照片。如果無法減少變色,則眼鏡應
be removed or the subject should be asked to use other glasses. See Figure D.27. A circled yellow “P” in the Figure indicates compliance depending on the acceptance policy of the issuer. 移除,或應要求受測者使用其他眼鏡。請參閱圖 D.27。圖中黃色圓圈的「P」表示符合發行者接受政策的規定。
Figure D. 27 - Examples for compliance of glasses 圖 D.27 — 眼鏡合規範例
Any lighting artefacts present on the region of the glasses shall not obscure eye details such that identification becomes difficult. Glasses may be repositioned to eliminate lighting artefacts, but frames shall not obscure eye details. The pitch shall not be increased by moving the head forward. 眼鏡區域上出現的任何光學偽影不應遮蔽眼睛細節,以免辨識困難。眼鏡可以重新調整位置以消除光學偽影,但鏡框不應遮蔽眼睛細節。不應透過將頭部向前移動來增加俯仰角。
Rims and frames of glasses shall not obscure the eyes as well as the EVZ. The irises of both eyes shall be visible to the same extent as without glasses. Frames should not be thicker than 5%5 \% of the IED (typically 3-4mm3-4 \mathrm{~mm} ). A subject wearing heavier frames should be asked to use other glasses or to remove their glasses. 眼鏡的邊緣和鏡框不應遮蔽眼睛以及眼睛可見區域(EVZ)。雙眼虹膜的清晰度應與未戴眼鏡時相同。鏡框的厚度不應超過瞳距(IED)的 5%5 \% (通常為 3-4mm3-4 \mathrm{~mm} )。應要求佩戴較厚鏡框的受試者使用其他眼鏡或取下眼鏡。
D.1.4.3.5 Accessories: Head coverings D.1.4.3.5 配件:頭部覆蓋物
The region of the face, from the crown to the base of the chin, and from ear-to-ear, shall be clearly visible. Special care shall be taken in cases when veils, scarves or head covering cannot be removed for religious reasons to ensure these coverings do not obscure any face features and do not generate shadow. Head coverings shall not be accepted except in circumstances specifically approved by the issuing state of the MRTD. Such circumstances may be religious, medical or cultural. If head coverings are allowed, they shall be firm fitting and of a plain uniform colour with no pattern and no visible perforations and the region between hair lines, both forwards of the ears and chin including cheeks, mouth, eyes, and eyebrows shall be visible without any distortion or shadows. For examples, see Figure D.28. 臉部區域,從頭頂到下巴底部,以及從一耳到另一耳,都應清晰可見。當因宗教原因無法移除面紗、圍巾或頭巾時,應特別注意確保這些遮蓋物不會遮擋任何臉部特徵,也不會產生陰影。除非經由旅行證件核發國家的特別批准,否則不應接受頭部遮蓋物。此類情況可能基於宗教、醫療或文化因素。如果允許佩戴頭部遮蓋物,則應緊密貼合,顏色應為素色且無圖案,無明顯穿孔,並且髮際線之間、耳朵前方和下巴(包括臉頰、嘴巴、眼睛和眉毛)的區域應清晰可見,無任何扭曲或陰影。範例請參閱圖 D.28。
The elliptically shaped region between the following face feature points as defined in ISO/IEC 14496-2 shall be visible without any intensive shadows: 依據 ISO/IEC 14496-2 定義,以下臉部特徵點之間的橢圓形區域應清晰可見,且無明顯陰影:
2.1: Bottom of the chin, 2.1:下巴底部,
10.9: Upper contact point between left ear and face, 10.9:左耳與臉部之間的上方接觸點,
10.10: Upper contact point between right ear and face, and 10.10:右耳與臉部之間的上方接觸點,以及
11.1: Middle border between hair and forehead. 11.1:頭髮與額頭之間的中間邊界。
An issuer may or may not require that the ears are visible. The capture process should minimize shadows and obscuration of features in the face region. This might involve adjustment of the head coverings. See Figure D.28. 發行者可能會或可能不會要求耳朵可見。擷取過程應盡量減少臉部區域特徵的陰影和遮蔽。這可能涉及調整頭部覆蓋物。請參閱圖 D.28。
a) Compliance depends on issuer policy a) 合規性取決於發行者政策
d) Non-uniform head covering d) 不均勻的頭部覆蓋物
b) Compliance depends on issuer policy b) 合規性取決於發行者政策
c) Face not completely visible c) 臉部未完全可見
f) Low background contrast f) 背景對比度低
Figure D. 28 - Head covering examples 圖 D.28 — 頭部覆蓋物範例
D.1.4.3.6 Accessories: Face ornamentation D.1.4.3.6 配件:臉部飾品
Face ornamentation which obscures the face shall not be present. Concerning face ornaments not obscuring the face, the issuer may use its discretion as to the extent to which face ornaments may appear in the face portrait. In any case, only permanently worn face ornaments may appear in the face portrait. See Figure D.29. 不得出現遮蔽臉部的臉部飾物。關於不遮蔽臉部的臉部飾物,發行者可自行決定臉部飾物在臉部肖像中出現的程度。無論如何,臉部肖像中只能出現永久佩戴的臉部飾物。請參閱圖 D.29。
Figure D. 29 - Face ornamentation examples 圖 D.29 – 臉部裝飾範例
People usually try to look better than normal in an ID photo. In some extreme cases an excessive use of make-up affects computerized as well as human face recognition capabilities. Therefore, the subject should only wear typical every day make-up. 人們通常會試圖在身分證照片中看起來比平常更好。在某些極端情況下,過度使用化妝品會影響電腦和人臉辨識能力。因此,受試者應只化日常妝。
There shall be no dirt visible on the face in a captured face portrait. It should be considered that dermatological problems could cause skin properties that look like dirt. The hair of the subject shall not cover any part of the eyes. The hair should not cover any part of the EVZ. See Figure D.30. Eye patches shall not be worn unless required for a medical reason. 在拍攝的臉部肖像中,臉部不得有可見的髒污。應考量皮膚病問題可能導致皮膚特性看起來像髒污。受試者的頭髮不得遮蓋眼睛的任何部分。頭髮不應遮蓋 EVZ 的任何部分。請參閱圖 D.30。除非有醫療原因,否則不得佩戴眼罩。
Figure D. 30 - Hair style examples 圖 D.30 - 髮型範例
D.1.4.4 Face portrait dimensions and head location D.1.4.4 臉部肖像尺寸與頭部位置
The head shall be centred in the final face portrait as described in this clause. The referenced feature points are defined in ISO/IEC 14496-2. See Figure 9, Figure D.32, and Table D.8. 頭部應如本條款所述,置於最終臉部肖像的中央。所參考的特徵點定義於 ISO/IEC 14496-2。請參閱圖 9、圖 D.32 和表 D.8。
The image width A to image height B aspect ratio should be between 74%74 \% and 80%80 \%. The imaginary line H is defined as the (almost horizontal) line through the eye centres of the left (feature point 12.1) and the right eye (feature point 12.2). 影像寬度 A 與影像高度 B 的長寬比應介於 74%74 \% 和 80%80 \% 之間。假想線 H 定義為穿過左眼(特徵點 12.1)和右眼(特徵點 12.2)中心(幾乎水平)的線。
The centre of H is the face midpoint M . The horizontal distance M_(h)\mathrm{M}_{\mathrm{h}} between the left image border and M shall be between 45%45 \% and 55%55 \% of A . The vertical distance M_(v)\mathrm{M}_{\mathrm{v}} between the top image border and M shall be between 30%30 \% and 50%50 \% of B . The mouth centre (feature point 2.3 ) and M define the imaginary (almost vertical) line V . Note, that V and H are not necessarily perpendicular. H 的中心是臉部中點 M。左側影像邊界與 M 之間的水平距離 M_(h)\mathrm{M}_{\mathrm{h}} 應介於 A 的 45%45 \% 和 55%55 \% 之間。頂部影像邊界與 M 之間的垂直距離 M_(v)\mathrm{M}_{\mathrm{v}} 應介於 B 的 30%30 \% 和 50%50 \% 之間。嘴巴中心(特徵點 2.3)和 M 定義了假想的(幾乎垂直的)線 V。請注意,V 和 H 不一定垂直。
The head width W is defined as the distance between the two imaginary lines parallel to the line V ; each imaginary line is drawn between the upper and lower lobes of each ear (feature points 10.2/10.6 for the right and 10.1//10.510.1 / 10.5 for the left ear). The W to A ratio shall be between 50%50 \% and 75%75 \%. This constraint is more important than including the entire hairline in the photograph for subjects with large hair volume. 頭部寬度 W 定義為兩條平行於線 V 的假想線之間的距離;每條假想線都繪製在每隻耳朵的上耳垂和下耳垂之間(右耳為特徵點 10.2/10.6,左耳為 10.1//10.510.1 / 10.5 )。W 與 A 的比例應介於 50%50 \% 和 75%75 \% 之間。對於髮量大的受試者,此限制比在照片中包含整個髮際線更重要。
The head length L is defined as the distance between the base of the chin (feature point 2.1) and the crown (feature point 11.4) measured on the imaginary line V. If these feature points are not exactly located at V , the vertical projection of them to V shall be used. The L to B ratio shall be between 60%60 \% and 90%90 \%. 頭部長度 L 定義為下巴底部(特徵點 2.1)與頭頂(特徵點 11.4)之間沿假想線 V 測量的距離。如果這些特徵點未精確位於 V 上,則應使用它們在 V 上的垂直投影。L 與 B 的比率應介於 60%60 \% 和 90%90 \% 之間。
Often, the location of crown, chin or ears cannot be determined precisely. In such a case, a good guess shall be made. 通常,頭頂、下巴或耳朵的位置無法精確判斷。在這種情況下,應進行合理猜測。
For examples see Figure D.31. 如需範例,請參閱圖 D.31。
Figure D. 31 - Face portrait dimensions and head location examples 圖 D.31 — 臉部肖像尺寸與頭部位置範例
Table D. 8 - Geometric face portrait requirements 表 D. 8 - 臉部幾何肖像要求
Line through mouth centre (feature point 2.3) and M 穿過嘴巴中心(特徵點 2.3)和 M 的線
W
Head width: Distance between the two imaginary lines parallel to the line V; each imaginary line is drawn between the upper and lower lobes of each ear (feature points 10.2/10.6 for the right and 10.1/10.5 for the left ear) 頭部寬度:兩條平行於 V 線的假想線之間的距離;每條假想線繪製於每隻耳朵的耳垂上下緣之間(右耳為特徵點 10.2/10.6,左耳為 10.1/10.5)
Head length: Distance between the base of the chin (feature point 2.1) and the crown (feature point 11.4) measured on the imaginary line V , if these feature points are not exactly located at V, the vertical projection of them to V shall be used 頭部長度:下巴底部(特徵點 2.1)與頭頂(特徵點 11.4)之間在假想線 V 上測量的距離,如果這些特徵點未精確位於 V 上,則應使用它們在 V 上的垂直投影。
Term Description Requirement
A Image width
B Image height 74% <= A//B <= 80%
H Line through the centres of the left (feature point 12.1) and the right eye (feature point 12.2)
M Face centre (midpoint of H)
M_(h) Distance from the left side of the image to M 45% <= M_(h)//A <= 55%
M_(v) Distance from the top of the image to M 30% <= M_(v)//B <= 50%
V Line through mouth centre (feature point 2.3) and M
W Head width: Distance between the two imaginary lines parallel to the line V; each imaginary line is drawn between the upper and lower lobes of each ear (feature points 10.2/10.6 for the right and 10.1/10.5 for the left ear) 50% <= W//A <= 75%
L Head length: Distance between the base of the chin (feature point 2.1) and the crown (feature point 11.4) measured on the imaginary line V , if these feature points are not exactly located at V, the vertical projection of them to V shall be used 60% <= L//B <= 90%| Term | Description | Requirement |
| :--- | :--- | :--- |
| A | Image width | |
| B | Image height | $74 \% \leq \mathrm{A} / \mathrm{B} \leq 80 \%$ |
| H | Line through the centres of the left (feature point 12.1) and the right eye (feature point 12.2) | |
| M | Face centre (midpoint of H) | |
| $\mathrm{M}_{\mathrm{h}}$ | Distance from the left side of the image to M | $45 \% \leq \mathrm{M}_{\mathrm{h}} / \mathrm{A} \leq 55 \%$ |
| $\mathrm{M}_{\mathrm{v}}$ | Distance from the top of the image to M | $30 \% \leq \mathrm{M}_{\mathrm{v}} / \mathrm{B} \leq 50 \%$ |
| V | Line through mouth centre (feature point 2.3) and M | |
| W | Head width: Distance between the two imaginary lines parallel to the line V; each imaginary line is drawn between the upper and lower lobes of each ear (feature points 10.2/10.6 for the right and 10.1/10.5 for the left ear) | $50 \% \leq \mathrm{W} / \mathrm{A} \leq 75 \%$ |
| L | Head length: Distance between the base of the chin (feature point 2.1) and the crown (feature point 11.4) measured on the imaginary line V , if these feature points are not exactly located at V, the vertical projection of them to V shall be used | $60 \% \leq \mathrm{L} / \mathrm{B} \leq 90 \%$ |
NOTE Both ISO/IEC 19794-5:2005 (edition 1) and ISO/IEC 19794-5:2011 (edition 2) comply with Table D. 8 geometric requirements for Full Frontal image type and Token Face image type. 註:ISO/IEC 19794-5:2005(第一版)和 ISO/IEC 19794-5:2011(第二版)均符合表 D.8 中全正面影像類型和標記臉部影像類型的幾何要求。
Figure D. 32 shows a typical example of a face portrait. In Figure D. 32 a) the intersection of the two rectangles marks the region where the centre point M shall be located. In Figure D. 32 b) the smaller rectangle in the face portrait shall be completely included in the head; the head itself shall be completely included in the larger rectangle. Note that the locations of these two rectangles do not depend on the location of M, the rectangles can be moved freely and independently from each other as long as they stay parallel to the borders of the image. Figure D. 33 gives samples where the faces do not fit into the larger rectangle or do not fill the smaller rectangle. 圖 D.32 顯示了臉部肖像的典型範例。在圖 D.32 a)中,兩個矩形的交點標示了中心點 M 應位於的區域。在圖 D.32 b)中,臉部肖像中的較小矩形應完全包含在頭部內;頭部本身應完全包含在較大矩形內。請注意,這兩個矩形的位置不取決於 M 的位置,只要它們保持與影像邊界平行,這些矩形就可以自由且獨立地移動。圖 D.33 提供了臉部不符合較大矩形或未填滿較小矩形的範例。
Figure D. 32 - Sample face portraits with the respective minimal and maximal head dimensions 圖 D. 32 - 臉部肖像範例,包含各自的最小和最大頭部尺寸
Figure D. 33 - Sample face portraits not complying with minimal and maximal head dimensions 圖 D. 33 - 不符合最小和最大頭部尺寸的臉部肖像範例
D.1.4.5 Children D.1.4.5 兒童
D.1.4.5.1 General D.1.4.5.1 概述
This subclause specifies additional guidance for capturing face portraits of children. Care should be taken to capture such images according to the specifications; however, sometimes this is not possible or would cause huge discomfort. Therefore, some requirements may be relaxed for children as specified below. See Figure D.34, Figure D. 35 and Figure D. 36 for sample images. 本子條款規定了拍攝兒童臉部肖像的額外指導。應注意根據規範拍攝此類影像;然而,有時這是不可能或會造成極大的不適。因此,對於兒童,某些要求可酌情放寬,如下所述。範例影像請參閱圖 D.34、圖 D.35 和圖 D.36。
D.1.4.5.2 Children below one year D.1.4.5.2 一歲以下兒童
Deviating from the specifications in D.1.4, babies under one year should be in an upright position, but it is acceptable to capture the face portrait with the baby lying on a white or plain light-coloured blanket which conforms to the requirements in D.1.4.2.5 and D.1.4.2.9. Alternatively, the baby may be placed in a baby seat as long as the background behind the head of the baby conforms to the requirements above and no portions of the baby seat are visible in the face portrait. 偏離 D.1.4 中的規範,一歲以下嬰兒應處於直立姿勢,但可接受嬰兒躺在符合 D.1.4.2.5 和 D.1.4.2.9 要求之白色或淺色素色毯子上拍攝臉部肖像。或者,嬰兒可放置於嬰兒座椅中,只要嬰兒頭部後方的背景符合上述要求,且嬰兒座椅的任何部分在臉部肖像中不可見。
Deviating from the specifications in D.1.4.3.3, it is not necessary that babies under one year have their eyes open. 偏離 D.1.4.3.3 中的規範,一歲以下嬰兒不需張開眼睛。
Hands, arms and other body parts of an assisting person used to support the positioning of the subject, e.g., parents supporting their child, shall not be visible in the image. Shadows of these assistant parts shall not be visible on the face portrait or in the background. 協助者用於支撐受測者定位的手、手臂及其他身體部位(例如:父母支撐其子女)不得出現在影像中。這些協助部位的陰影不得出現在臉部肖像或背景中。
D.1.4.5.3 Children below six years D.1.4.5.3 六歲以下兒童
Deviating from the specifications in D.1.4.3.1, children aged six and under shall face the camera within an angle of +-15^(@)\pm 15^{\circ} in pitch, yaw, and roll. Deviating from the specifications in D.1.4.3.2, children aged six and under do not need to have a neutral expression. For infants under the age of six, images are acceptable as long as the infant is awake, has his or her eyes open, there are no other people or objects in the photo and the background is uniform and face portrait meets the colour requirements in D.1.4.2.9. 與 D.1.4.3.1 中的規範不同,六歲及以下兒童應在俯仰、偏航和滾動角度為 +-15^(@)\pm 15^{\circ} 的範圍內面向攝影機。與 D.1.4.3.2 中的規範不同,六歲及以下兒童無需保持中性表情。對於六歲以下嬰兒,只要嬰兒清醒、眼睛睜開、照片中沒有其他人或物體,且背景均勻,臉部肖像符合 D.1.4.2.9 中的顏色要求,則影像可接受。
D.1.4.5.4 Children below eleven years D.1.4.5.4 十一歲以下兒童
Deviating from the specifications in D.1.4.4 for children of up to eleven years, L//B\mathrm{L} / \mathrm{B} shall be between 50%50 \% and 90%90 \%. Furthermore, M_(y)//B\mathrm{M}_{\mathrm{y}} / \mathrm{B} shall be between 30%30 \% and 60%60 \%. 偏離 D.1.4.4 中針對十一歲以下兒童的規範, L//B\mathrm{L} / \mathrm{B} 應介於 50%50 \% 和 90%90 \% 之間。此外, M_(y)//B\mathrm{M}_{\mathrm{y}} / \mathrm{B} 應介於 30%30 \% 和 60%60 \% 之間。
Figure D. 34 - Compliant child face portraits 圖 D. 34 - 符合規範的兒童臉部肖像
Figure D. 35 - Examples of additional objects visible in the image 圖 D.35 – 影像中可見的其他物件範例
a) Not looking into the camera a) 未直視攝影機
b) Eyes closed b) 眼睛閉合
c) Cap c) 蓋子
d) No neutral expression d) 無中性表情
Figure D. 36 - Examples of non-compliant poses and expressions 圖 D. 36 - 不合規的姿勢與表情範例
D.1.5 Image storage in the chip D.1.5 影像於晶片中的儲存
D.1.5.1 General D.1.5.1 概述
This subclause specifies properties of the face portrait to be electronically stored in a MRTD. Figure D. 37 shows the content of D.1.5 in the MRTD production process chain. 本子條款規定了應以電子方式儲存於機器可讀旅行證件(MRTD)中的臉部肖像屬性。圖 D. 37 顯示了 D.1.5 在 MRTD 製作流程鏈中的內容。
Figure D. 37 - Content of D.1.5 in the MRTD production process chain (boxed in red) 圖 D. 37 - MRTD 製作流程鏈中 D.1.5 的內容(紅框處)
The requirements and recommendations given in this subclause shall ensure that the photographic requirements given in D.1.4.2 are retained in the face portrait that is finally stored in a MRTD, with the exception of the pixel count. The minimal requirements specified in this subclause apply to the image finally stored in Data Group 2 as defined in ICAO Doc 9303. 本子條款所提出的要求與建議,應確保最終儲存於機器可讀旅行證件(MRTD)中的臉部肖像,除了像素數量外,皆符合 D.1.4.2 所列的攝影要求。本子條款所規定的最低要求適用於最終儲存於國際民航組織(ICAO)Doc 9303 所定義的資料組 2 中的影像。
ISO/IEC 39794-5:2019(E)
A submitted face portrait shall have been captured within the last six months before application. Face portraits with a capture time dating back more than three months should not be accepted. Issuers should consider the use of the metadata encoded with the digital image to assure that the photograph is recent. See Table D.9. 提交的臉部肖像應在申請前六個月內拍攝。拍攝時間超過三個月的臉部肖像不應被接受。發證機關應考慮使用數位影像中編碼的後設資料,以確保照片為近期拍攝。請參閱表 D.9。
Table D. 9 - Capture time requirements and recommendations 表 D.9 - 拍攝時間要求與建議
At most three months before application. 至多在申請前三個月。
"Criterion:
Capture time" Requirement At most six months before application.
Best practice At most three months before application.| Criterion: <br> Capture time | Requirement | At most six months before application. |
| :--- | :--- | :--- |
| | Best practice | At most three months before application. |
D.1.5.2 Data format D.1.5.2 資料格式
Face portraits of a subject to be stored in the MRTD chip 儲存於 MRTD 晶片中的人臉肖像
Shall be stored in one of the following formats: JPEG, JPEG2000, 應以以下其中一種格式儲存:JPEG、JPEG2000、
Should have a minimum IED of 90 pixels, preferably of 120 pixels (see Table D.10), 最小瞳距應為 90 像素,較佳為 120 像素(參見表 D.10),
Shall be in colour. 應為彩色。
These specifications provide adequate spatial sampling rate for use on the MRTD while maintaining an adequate quality for human and machine face recognition purposes. 這些規範提供了足夠的空間取樣率,可用於機器可讀旅行證件(MRTD),同時保持足夠的品質以供人臉和機器人臉辨識之用。
Table D. 10 - IED requirements and recommendations for the chip image 表 D. 10 – 晶片影像的 IED 要求與建議
The pixel count specified in D.1.4.2.4 applies to the originally captured face portrait and not to the images to be stored in a passport. The processing steps between capturing and passport production might lead to information losses. It is therefore recommended that a higher resolution version of the image is stored in the issuer’s repository. D.1.4.2.4 中指定的像素數適用於原始擷取的人臉肖像,而非儲存在護照中的影像。擷取與護照製作之間的處理步驟可能導致資訊遺失。因此,建議在發行者的程式碼儲存庫中儲存更高解析度的影像版本。
One of the three possible encodings shall be used: 應使用三種可能的編碼之一:
The JPEG sequential baseline (ISO/IEC 10918-1) mode of operation and encoded in the JFIF file format (the JPEG file format). JPEG 循序基準(ISO/IEC 10918-1)操作模式,並以 JFIF 檔案格式(JPEG 檔案格式)編碼。
The JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossy, and encoded in the JP2 file format (the JPEG2000 file format). JPEG-2000 第一部分碼流格式(ISO/IEC 15444-1),有損壓縮,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼。
The JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossless, and encoded in the JP2 file format (the JPEG2000 file format). JPEG-2000 第一部分碼流格式(ISO/IEC 15444-1),無損壓縮,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼。
The coordinate origin shall be at the upper left given by coordinate (0,0)(0,0) with positive entries from left to right (first dimension) and top to bottom (second dimension). 座標原點應位於左上角,由座標 (0,0)(0,0) 給定,正值從左到右(第一維度)和從上到下(第二維度)。
D.1.5.3 Property mask D.1.5.3 屬性遮罩
The positions of the Properties element in the data structure described in this document should be set for: 本文件中所述資料結構中「屬性」(Properties) 元素的定位應設定為:
(medical) dark glasses; (醫用)深色眼鏡;
head coverings; 頭部覆蓋物;
left and right eye patches; 左眼及右眼眼罩;
glasses; 眼鏡;
biometric absence (conditions which could impact landmark detection). 生物特徵缺失(可能影響地標偵測的狀況)。
Additionally, the Subject height element should be encoded. 此外,應編碼受測者身高元素。
processing RAW images into the target encoding (once) and/or; 將 RAW 影像處理成目標編碼(一次)和/或;
compression as described in D.1.5.5, 如 D.1.5.5 所述的壓縮,
shall be applied on the captured image to create the face portrait. Any processing shall maintain the requirements given in D.1.3 and D.1.4. Any processing shall render skin and hair colours realistically enough to allow straightforward human identification of the MRTD holder. The face images shall not be modified locally, e.g., for removal of scars, pimples or other skin impurities or to modify the shape or location of the nose, the eyes, the eyebrows or any other face landmarks. The image shall not be modified locally by editing clothes (e.g., a turban). 應將其應用於擷取的影像以建立臉部肖像。任何處理都應保持 D.1.3 和 D.1.4 中給出的要求。任何處理都應將皮膚和頭髮的顏色渲染得足夠真實,以便人類能夠直接辨識 MRTD 持有者。臉部影像不應進行局部修改,例如,去除疤痕、青春痘或其他皮膚雜質,或修改鼻子、眼睛、眉毛或任何其他臉部特徵的形狀或位置。影像不應透過編輯衣物(例如,頭巾)進行局部修改。
In particular, any image processing targeting at background removal shall not be implemented. If necessary, the MRTD issuing authority may remove or alter the background in the printed image later in the MRTD production process. 特別是,不應實施任何針對背景移除的影像處理。如有必要,MRTD 發行機構可在 MRTD 製作過程中稍後移除或更改列印影像中的背景。
D.1.5.5 Compression D.1.5.5 壓縮
Captured face portraits of a subject should not sacrifice image quality by overly compressing the image. 受測者臉部肖像的擷取不應因過度壓縮影像而犧牲影像品質。
For maximum effect in human and automated face recognition, the raw image or an image with limited compression should be retained. The JPEG compression ratio shall not exceed 15:115: 1. 為了在人臉辨識和自動化人臉辨識中達到最佳效果,應保留原始影像或壓縮程度有限的影像。JPEG 壓縮比不得超過 15:115: 1 。
EXAMPLE In many cases, such images have a size of at least 12 kBytes for JPEG and JPEG2000 for storage in the chip of an electronic MRTD. The upper limit is defined by the available storage space available on the chip and reading time requirements. 範例:在許多情況下,這類影像的大小至少為 12 KB(適用於 JPEG 和 JPEG2000),以便儲存在電子機器可讀取旅行證件(eMRTD)的晶片中。上限由晶片上可用的儲存空間和讀取時間要求決定。
The image used for the printing process on the MRTD and for storage in the chip will give better results if not compressed beyond a ratio creating visible artefacts on the image when viewed at 100%100 \% magnification - where a single pixel in an image file is displayed by a single pixel on a monitor or viewing device. This allows for electronic judging of whether an image is overly compressed. 用於 MRTD 印刷及晶片儲存的影像,若在 100%100 \% 倍放大檢視時(影像檔中的單一像素由螢幕或檢視裝置上的單一像素顯示),其壓縮比例未產生可見瑕疵,將能呈現更佳效果。這有助於電子判斷影像是否過度壓縮。
Lossy compressions can only be applied once per each of the following steps: 有損壓縮只能在下列每個步驟中套用一次:
one initial compression by the camera itself, 相機本身進行的首次壓縮,
one compression done by the photographer or citizen, and 攝影師或公民進行的壓縮,以及
one compression done by the issuer. 發行者進行的壓縮。
JPEG 2000 enables compressing to a target file size. If using JPEG compression must be done iteratively to the target file size while reverting back to the original image rather than successive compressions. JPEG 2000 能夠壓縮至目標檔案大小。如果使用 JPEG 壓縮,則必須反覆壓縮至目標檔案大小,同時回復到原始影像,而非連續壓縮。
ISO/IEC 39794-5:2019(E)
D. 2 General purpose face image D.2 一般用途臉部影像
D.2.1 General D.2.1 概述
This annex describes a profiled face image that meets minimal requirements to acquire an image for general face recognition usage. 本附錄說明了符合最低要求、可用於一般人臉辨識的臉部影像。
One of the following encodings shall be used: 應使用下列其中一種編碼:
The JPEG sequential baseline (ISO/IEC 10918-1) mode of operation and encoded in the JFIF file format (the JPEG file format) JPEG 循序基準(ISO/IEC 10918-1)操作模式,並以 JFIF 檔案格式(JPEG 檔案格式)編碼
The JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossy or lossless, and encoded in the JP2 file format (the JPEG2000 file format) JPEG-2000 Part-1 程式碼串流格式(ISO/IEC 15444-1),有損或無損,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼
The PNG specification (ISO/IEC 15948). PNG shall not be used in its interlaced mode and not for images that have been JPEG compressed before. PNG 規範 (ISO/IEC 15948)。PNG 不得使用其交錯模式,也不得用於先前已進行 JPEG 壓縮的影像。
Landmarks should be determined on images before compression is applied. Landmarks should be included in the record format if they have been accurately determined, thereby providing the option that these parameters do not have to be re-determined when the image is processed for face recognition tasks. The landmarks should be determined by computer-automated detection mechanisms followed by human validation. It is recommended to encode the following landmarks: the middle point of the eyes (12.1 and 12.2), the base of the nose (9.4, 9.5, and 9.15) and the upper lip of the mouth (8.4, 8.1 and 8.3). 應在套用壓縮之前,先在影像上確定特徵點。如果特徵點已準確確定,則應將其包含在記錄格式中,從而提供選項,使這些參數在影像處理用於臉部辨識任務時,無需重新確定。特徵點應透過電腦自動偵測機制,然後進行人工驗證來確定。建議編碼以下特徵點:眼睛中點(12.1 和 12.2)、鼻基部(9.4、9.5 和 9.15)以及嘴唇上緣(8.4、8.1 和 8.3)。
The 2D Image representation block shall be present. The value of the 2D face image kind shall be General purpose. 2D 影像表示區塊應存在。2D 臉部影像種類的值應為通用。
D.2.2 Image data compression requirements and recommendations D.2.2 影像資料壓縮要求與建議
Best practice on compression without a region of interest is: 在沒有感興趣區域的情況下,最佳壓縮實務為:
a) The compressed file size should not be smaller than 11 KB on average. a) 壓縮檔案大小平均不應小於 11 KB。
b) JPEG2000 should be preferred over JPEG. b) JPEG2000 應優先於 JPEG。
JPEG2000 can be used to implement region of interest (ROI) compression, as it is a technique specified in the JPEG2000 standard and well defined for JPEG2000 software libraries. JPEG2000 ROI encoding can be used to achieve smaller file sizes. JPEG2000 可用於實作感興趣區域 (ROI) 壓縮,因為這是一種在 JPEG2000 標準中指定且為 JPEG2000 軟體庫明確定義的技術。JPEG2000 ROI 編碼可用於實現更小的檔案大小。
The inner region of an image are the pixels having XX and YY coordinates with -1,5w <= X <= 1,5w-1,5 w \leq X \leq 1,5 w and -1,8W <= Y <= 1,8W-1,8 \mathrm{~W} \leq \mathrm{Y} \leq 1,8 \mathrm{~W} in the Cartesian coordinate system with the landmark Prn as origin, where W is the inter-eye distance. The outer region of an image consists of all pixels of the image that are not in the inner region of that image. 影像的內部區域是指在以地標 Prn 為原點的笛卡兒座標系統中,具有 XX 和 YY 座標以及 -1,5w <= X <= 1,5w-1,5 w \leq X \leq 1,5 w 和 -1,8W <= Y <= 1,8W-1,8 \mathrm{~W} \leq \mathrm{Y} \leq 1,8 \mathrm{~W} 的像素,其中 W 是雙眼距離。影像的外部區域包含影像中所有不屬於該影像內部區域的像素。
The inner region of a face image used for comparison can be compressed to a low ratio, while the outer region of the image is compressed to a higher ratio. The resulting image is smaller in size, but those parts of the image used for comparison retain high quality while the remainder of the image maintains their usefulness for visual inspection. A standard compliant JPEG2000 decoder with ROI support will decode an ROI image regardless of the location of ROI regions. 用於比對的人臉影像內部區域可以被壓縮成較低的壓縮比,而影像的外部區域則被壓縮成較高的壓縮比。如此產生的影像尺寸較小,但用於比對的影像部分仍能保持高品質,而影像的其餘部分則維持其用於視覺檢查的實用性。支援感興趣區域(ROI)的標準相容 JPEG2000 解碼器將會解碼 ROI 影像,無論 ROI 區域的位置為何。
The use of region of interest compression for situations where computer alignment is performed without human verification is not recommended. It is important to note that additional compression can be achieved by defining inner and outer regions that are based on the face area. 不建議在未經人工驗證的情況下,對執行電腦對齊的狀況使用感興趣區域壓縮。值得注意的是,透過定義基於臉部區域的內部和外部區域,可以實現額外的壓縮。
When derived from a 300 dpi image, an inner region can be defined as including the entire face from crown to chin and ear to ear. Best practice indicates that a compression ratio of 60:160: 1 using JPEG2000 preserves comparison performance. If a 50:150: 1 ratio is used for the inner region, 200:1200: 1 can be used on the outer region with an acceptable level of degradation for visual inspection purposes. For a colour, 當從 300 dpi 影像衍生時,內部區域可定義為包含從頭頂到下巴、從耳朵到耳朵的整個臉部。最佳實務指出,使用 JPEG2000 的壓縮比為 60:160: 1 可保留比對效能。如果內部區域使用 50:150: 1 的比例,則外部區域可使用 200:1200: 1 ,且在視覺檢查目的下,其降級程度可接受。對於彩色、
D.2.3 Scene requirements and recommendations D.2.3 場景要求與建議
D.2.3.1 Pose D.2.3.1 姿勢
Pose is known to strongly affect performance of automated face recognition systems. Thus, the frontal pose shall be used. Rotation of the head shall be less than +-5^(@)\pm 5^{\circ} from frontal in pitch and yaw. Pose variations that lead to an in-plane rotation of the head can be more easily compensated by automated face recognition systems. Therefore, the rotation of the head shall be less than +-8^(@)\pm 8^{\circ} from frontal in roll. 姿勢已知會嚴重影響自動臉部辨識系統的效能。因此,應採用正面姿勢。頭部在俯仰和偏擺方向的旋轉應小於正面 +-5^(@)\pm 5^{\circ} 。導致頭部平面內旋轉的姿勢變化,自動臉部辨識系統較易補償。因此,頭部在滾轉方向的旋轉應小於正面 +-8^(@)\pm 8^{\circ} 。
The best practice is that the rotation of the head should be less than +-5^(@)\pm 5^{\circ} from frontal in every direction, roll, pitch and yaw. The optimum height of the camera is at the subject’s eye-level. Height adjustment can be done by either using a height-adjustable stool or adjusting the tripod’s height. The subject should be instructed to look directly at the camera and to keep his or her head erect and shoulders square to the camera. 最佳實務是頭部在每個方向(滾轉、俯仰和偏擺)的旋轉都應小於正面 +-5^(@)\pm 5^{\circ} 。攝影機的最佳高度應與受測者的視線齊平。高度調整可透過使用可調高矮的凳子或調整三腳架的高度來完成。應指示受測者直視攝影機,並保持頭部挺直,肩膀與攝影機呈方形。
D.2.3.2 Expression D.2.3.2 表情
Expression is known to strongly affect the performance of automated face recognition systems. It is recommended that the Expression element is present. 已知表情會嚴重影響自動臉部辨識系統的效能。建議應包含表情元素。
The expression should be neutral (non-smiling) with both eyes open normally (i.e. not wide-open), and mouth closed (mouth is closed if the distance between landmark 2.2 and 2.3 is less than 50%50 \% of the distance of landmark 2.3 and 8.2). Every effort should be made to have the supplied images to comply with this specification. A smile with closed or open jaw, raised eyebrows, eyes looking away from the camera, squinting or frowning are not recommended. 表情應保持中性(不微笑),雙眼正常睜開(即非完全睜開),嘴巴閉合(若地標 2.2 與 2.3 之間的距離小於地標 2.3 與 8.2 之間距離的 50%50 \% ,則視為嘴巴閉合)。應盡力確保所提供的影像符合此規範。不建議出現閉合或張開嘴巴的微笑、眉毛上揚、眼睛未注視鏡頭、瞇眼或皺眉的狀況。
D.2.3.3 Shoulders D.2.3.3 肩膀
Shoulders shall be square on to the camera. Portrait style photographs where the subject is looking over one shoulder are not acceptable. 肩膀應正對鏡頭。受試者側身回頭的肖像照不被接受。
D.2.3.4 Backgrounds D.2.3.4 背景
The specification of a certain background is not normative for the creation of general purpose face images. A consideration of the background is important for computer-based face recognition because the first step in the computer face recognition process is the segmentation of the face from the background. 特定背景的規範對於建立通用人臉影像並非規範性。考量背景對於電腦化人臉辨識很重要,因為電腦人臉辨識過程的第一步是將人臉從背景中分割出來。
D.2.3.5 Subject and scene lighting D.2.3.5 主體與場景照明
Lighting shall be equally distributed on the face. There shall be no significant direction of the light from the point of view of the photographer. The ratio between the median intensity on a square region centred around Landmarks 5.3 and 5.4 with side length 20%20 \% of the inter-eye distance shall be between 0,5 and 2,0 . 臉部光線應均勻分佈。從攝影師的角度來看,光線不應有明顯的方向性。以地標 5.3 和 5.4 為中心,邊長為兩眼距離 20%20 \% 的正方形區域中,中位數強度之比應介於 0.5 到 2.0 之間。
D.2.3.6 Hot spots, specular reflections, and other lighting artefacts D.2.3.6 熱點、鏡面反射及其他照明偽影
Hot spots (i. e., bright regions that results from light shining directly on the face) shall be absent. These artefacts typically occur when a high intensity focused light source is used for illumination. Diffused lighting, multiple balanced sources or other lighting methods shall be used. 應無熱點(即光線直接照射臉部所造成的明亮區域)。這些偽影通常發生在高強度聚焦光源用於照明時。應使用漫射照明、多個平衡光源或其他照明方法。
There shall be no lighting artefacts or flash reflections on glasses. Lighting artefacts covering any region of the eyes shall not be present. This applies to any region in the polygon between landmarks 3.8, 3.2,3.123.2,3.12 and 3.4 for the right eye and between landmarks 3.11,3.1,3.73.11,3.1,3.7 and 3.3 for the left eye. 眼鏡上應無照明偽影或閃光反射。眼睛任何區域的照明偽影均不應存在。這適用於右眼地標 3.8、 3.2,3.123.2,3.12 和 3.4 之間以及左眼地標 3.11,3.1,3.73.11,3.1,3.7 和 3.3 之間多邊形內的任何區域。
D.2.3.7 Eye visibility and eye glasses D.2.3.7 眼睛可見度與眼鏡
The eye pupils and irises shall be visible. There should be no shadows in the eye-sockets due to the brow. In cases where pupils or irises are not visible the pupil or iris not visible element in the Properties element shall be true. 眼睛的瞳孔和虹膜應清晰可見。眉毛不應在眼窩處造成陰影。如果瞳孔或虹膜不可見,則「Properties」元素中的「pupil or iris not visible」元素應為真。
Eye patches shall not be worn. An exception applies when the subject asserts a need to retain the patch (e.g., a medical reason); in these cases, the left eye patch or the right eye patch element in the Properties element shall be true. 不得配戴眼罩。當受測者聲明需要保留眼罩時(例如,醫療原因),則適用例外情況;在這些情況下,屬性元素中的左眼罩或右眼罩元素應為真。
Hair should not cover any part of the eyes. It is recommended that hair should not cover landmarks 3.2, 3.8, 3.12 for the right eye and Landmarks 3.1, 3.7 and 3.11 for the left eye, as well as region above these points that measures 5 % of inter-eye distance. 頭髮不應遮蓋眼睛的任何部分。建議頭髮不應遮蓋右眼的標誌點 3.2、3.8、3.12,以及左眼的標誌點 3.1、3.7 和 3.11,以及這些點上方佔據眼距 5%的區域。
If the subject normally wears glasses, they may wear glasses if permitted for the intended application when their photograph is taken. 如果受測者平常有戴眼鏡,在拍照時,若預期應用程式允許,他們可以戴眼鏡。
Glasses should be clear and transparent. This requirement is intended to exclude dark or otherwise opaque glasses. Tinted glasses or sunglasses shall not be worn. An exception applies when the subject asserts a medical reason to retain tinted glasses; in these cases, the dark glasses element in the Properties element shall be true. 眼鏡應清晰透明。此要求旨在排除深色或其他不透明的眼鏡。不得佩戴有色眼鏡或太陽眼鏡。當受測者聲稱有醫療原因需保留有色眼鏡時,則適用例外情況;在這些情況下,「屬性」元素中的「深色眼鏡」元素應為真。
If glasses are worn that tint automatically under illumination, they should be photographed without tint by tuning the direct illumination or background lighting. In cases where the tint cannot be reduced, the glasses shall be removed unless the subject asserts a medical reason to retain the glasses. In cases where tinted glasses are worn, the specification of dark glasses in the Properties element is recommended. 如果佩戴的眼鏡在光線下會自動變色,則應透過調整直接照明或背景照明,在無變色狀態下進行拍攝。如果無法減少變色,則應取下眼鏡,除非受測者聲稱有醫療原因需保留眼鏡。在佩戴有色眼鏡的情況下,建議在「屬性」元素中指定為深色眼鏡。
The frames of glasses shall not obscure the eyes. The frames shall not be thicker than 5%5 \% of the intereye distance. Rims of glasses are covering part of the eye if any part of rims covers any part of the area enclosed by landmarks 3.2,3.4,3.83.2,3.4,3.8 and 3.12 for the right eye and landmarks 3.1,3.3,3.73.1,3.3,3.7 and 3.11 for the left eye, as well as region around these points that measures 5%5 \% of inter-eye distance. If rims of glasses are not visible or are completely transparent, it is assumed that they do not cover any part of the eye. 眼鏡框不得遮擋眼睛。鏡框厚度不得超過眼距的 5%5 \% 。如果鏡框的任何部分遮蓋了右眼由標記點 3.2,3.4,3.83.2,3.4,3.8 和 3.12 所圍繞的區域,以及左眼由標記點 3.1,3.3,3.73.1,3.3,3.7 和 3.11 所圍繞的區域,以及這些點周圍測量為眼距 5%5 \% 的區域,則視為鏡框遮蓋了眼睛的一部分。如果鏡框不可見或完全透明,則假定它們沒有遮蓋眼睛的任何部分。
Lighting artefacts can typically be avoided by increasing the angle between the lighting, subject and camera to 45^(@)45^{\circ} or more. 通常可以透過將照明、受測者和攝影機之間的角度增加到 45^(@)45^{\circ} 或更大來避免光線偽影。
D.2.3.8 Head coverings D.2.3.8 頭部覆蓋物
In cases where head coverings are present the head coverings element in the Properties element shall be true. 若有頭部覆蓋物,則「屬性」元素中的「頭部覆蓋物」元素應為真。
Head coverings and shadows should be absent. An exception applies to cases in which a subject cannot remove a headdress, veil or scarf (e.g, for religious reasons). In such cases the capture process should minimize shadows and obscuration of the face features in the face region. This might involve adjustment of the head coverings. 頭部覆蓋物和陰影應避免。但若受測者因故(例如宗教因素)無法移除頭飾、面紗或圍巾,則為例外。在此情況下,擷取過程應盡量減少陰影及臉部區域特徵的遮蔽。這可能需要調整頭部覆蓋物。
D.2.4 Photographic requirements and recommendations D.2.4 攝影要求與建議
D.2.4.1 Purpose D.2.4.1 目的
Rather than impose a particular hardware and lighting capture system, this subclause specifies the desired output image properties. The requirements and recommendations apply to film as well as to digital photography. 本子條款旨在規定所需的輸出影像特性,而非強制採用特定的硬體和照明擷取系統。這些要求和建議適用於底片攝影和數位攝影。
This subclause describes the minimum relative dimensions of the full image with respect to the face. The requirements can be met by images taken in both face portrait and landscape mode, and Figure 9 shows a face portrait image and head outline to display lines H and V and dimensions A,B,W\mathrm{A}, \mathrm{B}, \mathrm{W}, and L which are referenced in the subclauses below. In addition to the requirements in D.2.4.2_\underline{\mathrm{D} .2 .4 .2} through D.2.4.14, the face shall be entirely visible from crown to chin and ear to ear in the image. 本子條款描述了完整影像相對於臉部的最小相對尺寸。這些要求可透過臉部肖像模式和橫向模式拍攝的影像來滿足,圖 9 顯示了臉部肖像影像和頭部輪廓,以顯示線條 H 和 V 以及尺寸 A,B,W\mathrm{A}, \mathrm{B}, \mathrm{W} 和 L,這些尺寸在以下子條款中有所提及。除了 D.2.4.2_\underline{\mathrm{D} .2 .4 .2} 到 D.2.4.14 中的要求外,臉部在影像中應從頭頂到下巴、從耳朵到耳朵完全可見。
NOTE For digital images the requirements related to the minimum inter-eye distance impose further requirements on the minimum head size. 註:對於數位影像,與最小眼距相關的要求對最小頭部尺寸提出了進一步的要求。
D.2.4.2 Contrast and saturation D.2.4.2 對比度與飽和度
For each patch of skin on the capture subject’s face, the gradations in textures shall be clearly visible, i.e., being of reasonable contrast. In this sense, there will be no saturation (over or under exposure) on the face. 對於擷取對象臉部的每個皮膚區塊,紋理的漸變應清晰可見,即具有合理的對比度。從這個意義上說,臉部不會出現飽和(過度曝光或曝光不足)現象。
The colour saturation of a 24 -bit colour image should be such that after conversion to greyscale, there are 7 bits of intensity variation in the face region of the image. 24 位元彩色影像的色彩飽和度應使其在轉換為灰階後,影像臉部區域的強度變化達到 7 位元。
D.2.4.3 Focus and depth of field D.2.4.3 對焦與景深
The subject’s captured image shall always be in focus from nose to ears and chin to crown. Although this may result in the background behind the subject being out of focus, this is not a problem. 受測者的擷取影像應始終保持從鼻子到耳朵、從下巴到頭頂都清晰對焦。儘管這可能導致受測者後方的背景失焦,但這並不是問題。
All images shall have sufficient depth of focus to maintain visibility of all of the subject’s face features greater than one milimetre in size (at the face) at time of capture. This is considered accomplished, if, e.g., the individual millimetre markings of rulers placed on the subject’s nose and ear facing the camera can be seen simultaneously in a captured test image. 所有影像應具有足夠的景深,以在擷取時保持受測者臉部所有大於一毫米(在臉部)的特徵可見。如果例如,在擷取測試影像中,同時可以看到放置在受測者鼻子和耳朵上、面向攝影機的尺子上的個別毫米標記,則認為已達成此要求。
In a typical photographic situation, for optimum quality of the captured face, the f-stop of the lens should be set at two (or more) f-stops below the maximum aperture opening when possible to obtain enough depth of field. 在典型的攝影情況下,為了獲得最佳的臉部擷取品質,鏡頭的光圈值應設定在最大光圈開口以下兩級(或更多)光圈值,以便在可能的情況下獲得足夠的景深。
If the camera lacks auto focus all subject positions will need to be maintained in a defined area for all image captures. 如果攝影機缺乏自動對焦功能,則所有受測者位置在所有影像擷取過程中,都需維持在一個定義的區域內。
D.2.4.4 Greyscale density D.2.4.4 灰階密度
The dynamic range of the image should have at least 7 bits of intensity variation (span a range of at least 128 unique values) in the face region of the image. The face region is defined as the region from crown to chin and from the left ear to the right ear. This recommendation may require camera, video digitizer or scanner settings to be changed on an individual basis when the skin tone is excessively lighter or darker than the average (present) population. 影像的動態範圍應在臉部區域具有至少 7 位元的強度變化(涵蓋至少 128 個獨特值的範圍)。臉部區域定義為從頭頂到下巴,以及從左耳到右耳的區域。當膚色比平均(現有)人口過度偏亮或偏暗時,此建議可能需要個別調整相機、視訊數位轉換器或掃描器的設定。
D.2.4.5 Unnatural colour D.2.4.5 不自然的顏色
Unnaturally coloured lighting (yellow, red, etc.) is not allowed. Care shall be taken to correct the white balance of image capture devices. The lighting shall produce a face image with natural looking flesh tones when viewed in typical examination environments. Images showing a red eye effect, i.e., the common appearance of red eyes on photographs taken with a photographic flash when the flash is too close to the lens, are not acceptable. The iris and the iris colour shall be visible. 不允許使用不自然顏色的照明(黃色、紅色等)。應注意校正影像擷取裝置的白平衡。照明應在典型的檢查環境中產生具有自然膚色的臉部影像。顯示紅眼效果的影像,即當閃光燈離鏡頭太近時,使用攝影閃光燈拍攝的照片中常見的紅眼現象,是不可接受的。虹膜和虹膜顏色應清晰可見。
Greyscale photographs should be produced from common incandescent light sources. Colour photographs should use colour-balancing techniques such as using high colour-temperature flash with standard film or tungsten-balanced film with incandescent lighting. 灰階照片應使用常見的白熾燈光源拍攝。彩色照片應使用色彩平衡技術,例如使用高色溫閃光燈搭配標準底片,或使用鎢絲燈平衡底片搭配白熾燈照明。
D.2.4.6 Colour or greyscale enhancement D.2.4.6 彩色或灰階增強
A process that overexposes or under-develops a colour or greyscale image for purposes of beauty enhancement or artistic pleasure is not allowed. The full spectrum shall be represented on the face 不允許為了美化或藝術目的而對彩色或灰階影像進行過度曝光或顯影不足的處理。臉部影像應適當地呈現完整的光譜。
ISO/IEC 39794-5:2019(E)
image where appropriate. Teeth and whites of eyes shall be clearly light or white (when appropriate) and dark hair or features (when appropriate) shall be clearly dark. 影像中,牙齒和眼白應清晰地呈現淺色或白色(適當時),而深色頭髮或特徵(適當時)應清晰地呈現深色。
D.2.4.7 Colour calibration D.2.4.7 色彩校正
Colour calibration using an 18%18 \% grey background or other method (such as white balancing) is recommended. 建議使用 18%18 \% 灰色背景或其他方法(例如白平衡)進行色彩校準。
D.2.4.8 Radial distortion of the camera lens D.2.4.8 相機鏡頭的徑向畸變
The fish eye effect associated with wide angle lenses which can result in the subject appearing to have an unusually large nose in the image shall not be present. 不應出現與廣角鏡頭相關的魚眼效果,該效果可能導致影像中的主體鼻子異常大。
While some distortion is almost always present during face portrait photography, the distortion should not be noticeable by human examination. 在人臉肖像攝影中,雖然幾乎總是存在一些失真,但這種失真不應被人眼察覺。
D.2.4.9 Horizontally centred face D.2.4.9 水平置中人臉
The approximate horizontal midpoints of the mouth and of the bridge of the nose define the imaginary line V (usually the symmetry axis of the face). Furthermore, the imaginary line H is defined as the line through the centres of the left and the right eye. The intersection of V and H defines the point M as the centre of the face. The X-coordinate M_(x)M_{x} of MM shall be between 45%45 \% and 55%55 \% of the image width. 嘴巴和鼻樑的近似水平中點定義了假想線 V(通常是臉部的對稱軸)。此外,假想線 H 定義為穿過左眼和右眼中心的線。V 和 H 的交點定義了點 M 作為臉部的中心。 M_(x)M_{x} 的 X 座標 MM 應介於影像寬度的 45%45 \% 和 55%55 \% 之間。
D.2.4.10 Vertical position of the face D.2.4.10 臉部的垂直位置
The Y-coordinate M_(y)\mathrm{M}_{\mathrm{y}} of M shall be between 30%30 \% and 50%50 \% of the image height. A single exception is allowed for children under the age of 11 years, in which case the higher limit shall be modified to 60%60 \% (i. e., the centre point of the head is allowed to be lower in the image for children under the age of 11). The origin 0 of the coordinate system is defined to be in the upper left corner of the image. M 的 Y 座標 M_(y)\mathrm{M}_{\mathrm{y}} 應介於影像高度的 30%30 \% 和 50%50 \% 之間。11 歲以下的兒童允許單一例外,在這種情況下,上限應修改為 60%60 \% (即,11 歲以下兒童的頭部中心點在影像中可較低)。座標系統的原點 0 定義為影像的左上角。
D.2.4.11 Width of the image D.2.4.11 影像寬度
To ensure that the entire face is visible in the image, the IED shall be between 25%25 \% and 37,5 % of the image width A. 為確保影像中可見整個臉部,IED 應介於影像寬度 A 的 25%25 \% 和 37.5% 之間。
D.2.4.12 Height of the image D.2.4.12 影像高度
In order to assure that the entire face is visible in the image, the minimum image height shall be specified by requiring that the eye-to-mouth distance (segment between M and (feature point 2.3 from ISO/IEC 14496-2:2004, Annex C) of the image shall be between 20%20 \% and 30%30 \% of the vertical height of the image B. A single exception is allowed for children under the age of 11 years, in which case the lower limit shall be modified to 15%15 \%. 為確保影像中能看見整張臉,應規定最小影像高度,要求影像中眼睛到嘴巴的距離(ISO/IEC 14496-2:2004 附錄 C 中的特徵點 2.3 與 M 之間的區段)應介於影像 B 垂直高度的 20%20 \% 和 30%30 \% 之間。11 歲以下的兒童允許一個例外,在此情況下,下限應修改為 15%15 \% 。
D.2.4.13 Image aspect ratio D.2.4.13 影像長寬比
The (image width: image height) aspect ratio should be between 1:1,251: 1,25 and 1:1,341: 1,34. 影像的長寬比(影像寬度:影像高度)應介於 1:1,251: 1,25 和 1:1,341: 1,34 之間。
D.2.4.14 Summary of photographic requirements D.2.4.14 攝影要求摘要
Table D. 11 below summarizes the photographic requirements for general purpose face images. 下表 D. 11 總結了通用人臉影像的攝影要求。
Table D. 11 - Summary of photographic requirements for general purpose face images 表 D. 11 — 通用人臉影像的攝影要求摘要
Clause Definition Requirements
D.2.4.1 General requirement Head entirely visible in the image
D.2.4.9 Horizontal position of the face 0,45A <= M_(x) <= 0,55A
D.2.4.10 Vertical position of the face 0,3B <= M_(y) <= 0,5B| Clause | Definition | Requirements |
| :--- | :--- | :--- |
| D.2.4.1 | General requirement | Head entirely visible in the image |
| D.2.4.9 | Horizontal position of the face | $0,45 \mathrm{~A} \leq \mathrm{M}_{\mathrm{x}} \leq 0,55 \mathrm{~A}$ |
| D.2.4.10 | Vertical position of the face | $0,3 \mathrm{~B} \leq \mathrm{M}_{\mathrm{y}} \leq 0,5 \mathrm{~B}$ |
Table D. 11 (continued) 表 D. 11(續)
Clause 條款
Definition 定義
Requirements 要求
D.2.4.10
Vertical position of the face (children below 11) 臉部的垂直位置(11 歲以下兒童)
Clause Definition Requirements
D.2.4.10 Vertical position of the face (children below 11) 0,3B <= M_(y) <= 0,6B
D.2.4.11 Width of head 0,25A <= IED <= 0,375A
D.2.4.12 Length of head 0,2B <= EMD <= 0,3B
D.2.4.12 Length of head (children below 11) 0,15B <= EMD <= 0,3B| Clause | Definition | Requirements |
| :--- | :--- | :--- |
| D.2.4.10 | Vertical position of the face (children below 11) | $0,3 \mathrm{~B} \leq \mathrm{M}_{\mathrm{y}} \leq 0,6 \mathrm{~B}$ |
| D.2.4.11 | Width of head | $0,25 \mathrm{~A} \leq \mathrm{IED} \leq 0,375 \mathrm{~A}$ |
| D.2.4.12 | Length of head | $0,2 \mathrm{~B} \leq \mathrm{EMD} \leq 0,3 \mathrm{~B}$ |
| D.2.4.12 | Length of head (children below 11) | $0,15 \mathrm{~B} \leq \mathrm{EMD} \leq 0,3 \mathrm{~B}$ |
D.2.5 Digital requirements and recommendations D.2.5 數位要求與建議
D.2.5.1 Geometry D.2.5.1 幾何
Digital cameras and scanners used to capture face images shall produce images with a pixel aspect ratio of 1:11: 1. That is, the number of pixels per inch in the vertical dimension shall equal the number of pixels per inch in the horizontal direction. 用於擷取臉部影像的數位相機和掃描器應產生像素長寬比為 1:11: 1 的影像。亦即,垂直方向的每英吋像素數應等於水平方向的每英吋像素數。
The origin of coordinates shall be at the upper left given by coordinate (0,0)(0,0) with positive entries from left to right (first dimension) and top to bottom (second dimension). 座標原點應位於左上角,由座標 (0,0)(0,0) 給定,其正值項目從左到右(第一維度)和從上到下(第二維度)排列。
D.2.5.2 Colour profile D.2.5.2 色彩設定檔
General purpose face images shall be represented as one of the following: 通用人臉影像應以下列其中一種方式呈現:
a) 24-bit RGB colour space where for every pixel, eight (8) bits will be used to represent each of the Red, Green, and Blue components. a) 24 位元 RGB 色彩空間,其中每個像素的紅、綠、藍元件將使用八 (8) 位元表示。
b) 8-bit monochrome colour space where for every pixel, (8) bits will be used to represent the luminance component. b) 8 位元單色色彩空間,其中每個像素的亮度元件將使用八 (8) 位元表示。
c) YUV422 colour space where twice as many bits are dedicated to luminance as to each of the two colour components. YUV422 images typically contain two 8 -bit Y samples along with one 8 -bit sample of each of UU and V in every four bytes. c) YUV422 色彩空間,其中分配給亮度(luminance)的位元數是分配給兩個色彩分量(colour components)中每個分量的兩倍。YUV422 影像通常每四個位元組包含兩個 8 位元 Y 樣本,以及一個 8 位元 U 和一個 8 位元 V 樣本。
Interlaced video frames are not allowed for the general purpose face image kind. All interlacing shall be absent (not removed, but absent). 一般用途臉部影像不允許使用交錯式視訊影格。所有交錯都應不存在(不是移除,而是不存在)。
D.2.5.3 Use of near infrared cameras D.2.5.3 近紅外線攝影機的使用
Dedicated near-infrared cameras shall not be used for acquisition of image of the general purpose face image kind. 專用近紅外線攝影機不得用於擷取一般用途臉部影像。
D.2.5.4 Pixel count D.2.5.4 像素計數
For an image for optimal human examination and permanent storage, the head width shall be at least 180 pixels, or roughly 90 pixels from eye centre to eye centre. 為了最佳的人工檢視和永久儲存影像,頭部寬度應至少為 180 像素,或約為兩眼中心間距 90 像素。
D.2.5.5 Post acquisition processing D.2.5.5 後期處理
No other post processing than in-plane rotation and/or cropping and/or down sampling and/or multiple compressions shall be applied to derive a general purpose face image from a captured image. Multiple (i.e., repeated) compressions should be avoided when generating general purpose face images. 除了平面內旋轉、裁切、縮小取樣或多次壓縮外,不應對擷取到的影像進行其他後處理,以產生通用臉部影像。在產生通用臉部影像時,應避免多次(即重複)壓縮。
D. 3 3D Textured face image D.3 3D 紋理臉部影像
D.3.1 General D.3.1 概述
This annex contains the description of an application profile for a 3D textured face image that meets the requirements of this document to acquire an image for 3 D face recognition. 本附錄包含一個 3D 紋理臉部影像應用程式個人資料的描述,該個人資料符合本文件要求,可用於擷取影像以進行 3D 臉部辨識。
The purpose of the 3D textured application profile is to encode the shape and the texture of the face (see Figure D.38) with high precision. For some use cases, the texture of the face is optional. By an optical backward projection of the 3D presentation to a virtual camera with defined lighting, similar image quality of the skin rendering compared to a 2D representation data with different viewing angles should be obtained. 3D 紋理應用程式個人資料的目的是以高精度編碼臉部的形狀和紋理(參見圖 D.38)。對於某些使用案例,臉部紋理是可選的。透過將 3D 呈現光學反向投影到具有定義照明的虛擬攝影機,應能獲得與具有不同視角的 2D 表示資料相似的皮膚渲染影像品質。
The 3D shape representation block shall be present. The value of the 3D face image kind shall be textured face image 3D. 3D 形狀表示區塊應存在。3D 臉部影像種類的值應為紋理臉部影像 3D。
The 3D representation data element shall contain only the vertex representation, that means range image and 3D point map shall not be used. 3D 表示資料元素應僅包含頂點表示,這表示不應使用範圍影像和 3D 點圖。
Figure D. 38 - Example of a 3D textured image representation data which is composed of the 3D data representation (a list of triangles consisting of 3 vertices) and a Texture map data 圖 D. 38 - 3D 紋理影像表示資料範例,其由 3D 資料表示(由 3 個頂點組成的三角形列表)和紋理貼圖資料組成
Each vertex of the 3D representation data has a 2D UV spatial reference to the Texture data. The data defined by the textured vertex representation is encoded in two different data structures: 3D 表示資料的每個頂點都具有指向紋理資料的 2D UV 空間參考。由紋理頂點表示定義的資料以兩種不同的資料結構編碼:
The mandatory 3D representation data for the shape of the face, where the 3D representation data is defined by a set of: 臉部形狀的強制性 3D 表示資料,其中 3D 表示資料由以下集合定義:
a) a Vertex block defined by the 3D coordinates X, Y, and Z and by the spatial coordinates U and V which refer to the Texture map, and a) 由 3D 座標 X、Y 和 Z 以及指向紋理貼圖的空間座標 U 和 V 定義的頂點區塊,以及
b) a Vertex triangle data block referring to the ordered list of vertices. b) 頂點三角形資料區塊,其參照頂點的有序列表。
The optional Texture map data for the texture image of the face, where one of the following possible encodings for the texture map data (which is a 2D image) shall be used: 臉部紋理影像的選用紋理貼圖資料,其中應使用下列其中一種可能的紋理貼圖資料(其為 2D 影像)編碼:
a) JPEG sequential baseline (ISO/IEC 10918-1) mode of operation and encoded in the JFIF file format (the JPEG file format), a) JPEG 循序基準(ISO/IEC 10918-1)操作模式,並以 JFIF 檔案格式(JPEG 檔案格式)編碼,
b) JPEG-2000 Part-1 code stream format (ISO/IEC 15444-1), lossy or lossless, and encoded in the JP2 file format (the JPEG2000 file format), or b) JPEG-2000 第一部分碼流格式(ISO/IEC 15444-1),有損或無損,並以 JP2 檔案格式(JPEG2000 檔案格式)編碼,或
c) PNG specification (ISO/IEC 15948). PNG shall not be used in its interlaced mode and not for images that have been JPEG compressed before. c) PNG 規範 (ISO/IEC 15948)。PNG 不得使用其交錯模式,也不得用於先前已進行 JPEG 壓縮的影像。
The formats unknown and other shall not be used. The specification of the Texture map image is stored inside of the image header, i.e., the image width, image height, channel numbers, number of bits per channel and ICC profile. 不得使用未知及其他格式。紋理貼圖影像的規格儲存於影像標頭內,即影像寬度、影像高度、通道數、每個通道的位元數及 ICC 設定檔。
Landmarks shall be defined with 3D coordinates. 地標應以 3D 座標定義。
Landmarks should be determined on images before compression is applied. Landmarks should be included in the record format if they have been accurately determined, thereby providing the option that these parameters do not have to be re-determined when the image is processed for face recognition tasks. The landmarks may be determined by computer-automated detection mechanisms. If necessary, a human validation can be applied in such case. It is recommended to add the following landmarks to the encoding of a 3D image: 地標應在套用壓縮前於影像上確定。如果地標已精確確定,則應將其包含在記錄格式中,從而提供在處理影像以進行臉部辨識任務時無需重新確定這些參數的選項。地標可由電腦自動偵測機制確定。如有必要,在此情況下可進行人工驗證。建議將以下地標新增至 3D 影像的編碼中:
the eye centres (12.1 and 12.2), 眼睛中心(12.1 和 12.2),
the base of the nose (9.4, 9.5, and 9.15), and 鼻子的底部(9.4、9.5 和 9.15),以及
the upper lip of the mouth (8.4, 8.1 and 8.3). 嘴巴的上唇(8.4、8.1 和 8.3)。
D.3.2 Image data compression requirements and recommendations D.3.2 影像資料壓縮要求與建議
Best practice on compression without region of interest is: 無感興趣區域壓縮的最佳實踐為:
a) The compressed file size should not be smaller than 100 KB on average. a) 壓縮檔案大小平均不應小於 100 KB。
b) JPEG2000 should be preferred over JPEG. b) 應優先使用 JPEG2000 而非 JPEG。
NOTE For the Textured 3D representation, the texture map data refer only to the face texture description and not, e.g., to the background or the shoulders. As a consequence, there is no significant gain to use the implement region of interest (ROI) compression of JPEG2000. 註:對於紋理 3D 表示,紋理貼圖資料僅指臉部紋理描述,而非例如背景或肩膀。因此,使用 JPEG2000 的感興趣區域 (ROI) 壓縮並無顯著效益。
D.3.3 Scene requirements and recommendations D.3.3 場景要求與建議
D.3.3.1 Pose of the 3D representation D.3.3.1 3D 表徵的姿態
Pose is known to strongly affect performance of automated face recognition systems. However, this sensibility is less important for 3D representation which can be rotated without losing information after acquisition. 已知姿態會嚴重影響自動人臉辨識系統的效能。然而,對於 3D 表徵而言,這種敏感度較不重要,因為 3D 表徵在擷取後可以旋轉而不會遺失資訊。
For the pose of the Textures 3D face image representation the following requirements on the subject position, and on the 3D acquisition system geometry apply: 對於 Textures 3D 臉部影像表徵的姿態,適用以下關於受試者位置和 3D 擷取系統幾何的要求:
Subject position: 受試者位置:
The pose of the subject shall be with the head in the rest position. The eyes shall be looking straight forward according to the horizontal axis. The with shoulder position shall be perpendicular to the gaze axis. 受測者的姿勢應為頭部處於靜止位置。眼睛應沿水平軸線直視前方。肩膀位置應垂直於凝視軸線。
Rotation of the head shall be less than +-5^(@)\pm 5^{\circ} in pitch, yaw and roll. 頭部在俯仰、偏擺和滾轉方向的旋轉應小於 +-5^(@)\pm 5^{\circ} 。
3D acquisition system geometry: 3D 擷取系統幾何:
The optical axis of the 3D acquisition system shall be horizontal, at the same height of the eye subject, and perpendicular to the support line passing by the two eyes. Height adjustment may be done by either using a height-adjustable stool or by adjusting the acquisition system height. 3D 擷取系統的光軸應為水平,與受測者眼睛高度相同,並垂直於穿過雙眼的支撐線。高度調整可透過使用可調整高度的凳子或調整擷取系統的高度來完成。
The subject should be instructed to orient his or her gaze paralleled to the optical axis via a visible sign. The subject should be instructed to keep his or her head erect and shoulders square to the 3D acquisition system. 應指示受測者透過可見標誌將其視線與光軸平行。應指示受測者保持頭部挺直,肩膀與 3D 擷取系統呈直角。
The subject shall not move during the acquisition. In particular, no instruction shall request a subject movement during capturing. 受測者在擷取期間不得移動。特別是,在擷取期間不應有任何指示要求受測者移動。
ISO/IEC 39794-5:2019(E)
D.3.3.2 Expression D.3.3.2 表情
Expression is known to strongly affect the performance of automated face recognition systems. It is recommended that the Expression element is present. 已知表情會嚴重影響自動臉部辨識系統的效能。建議應包含表情元素。
The expression should be neutral (non-smiling) with both eyes open normally (i.e., not wide-open), and mouth closed (the mouth is closed if the distance between landmarks 2.2 and 2.3 is less than 50%50 \% of the distance of landmarks 2.3 and 8.2.) Every effort should be made supply images complying with this specification. To smile with closed or open jaw is not recommended, neither are raised eyebrows, eyes looking away from camera, squinting or frowning. 表情應保持中性(不微笑),雙眼正常睜開(即,非完全睜開),嘴巴閉合(若地標 2.2 與 2.3 之間的距離小於地標 2.3 與 8.2 之間距離的 50%50 \% ,則視為嘴巴閉合)。應盡一切努力提供符合此規範的影像。不建議閉合或張開下顎的微笑,也不建議挑眉、眼睛未注視鏡頭、瞇眼或皺眉。
The 3D textured face image acquisition system shall not allow expression change. The 3D face image acquisition process should be fast enough to ensure that expression does not change. Morphing and interpolation after the acquisition should not change the expression. 3D 紋理臉部影像擷取系統不應允許表情變化。3D 臉部影像擷取過程應足夠快速,以確保表情不變。擷取後的變形和插值不應改變表情。
D.3.3.3 Shoulders D.3.3.3 肩膀
Shoulders shall be square on to the camera. Portrait style photographs where the subject is looking over one shoulder are not acceptable. 肩膀應正對鏡頭。受試者側身回頭的肖像照不被接受。
D.3.3.4 Background D.3.3.4 背景
The background shall not be stored inside the 3D vertex encoding. A 3D acquisition system shall be able to differentiate the background from the face, based on the depth data along the ZZ axis. 背景不應儲存在 3D 頂點編碼中。3D 擷取系統應能根據沿著 ZZ 軸的深度資料,區分背景和臉部。
The minimal distance between face and background should be 400 mm . 臉部與背景之間的最小距離應為 400 公釐。
The colour of the background should be uniform and should have a contrast to skin and hairs. 背景顏色應均勻,並與皮膚和頭髮形成對比。
NOTE Some configurations of the background can improve the segmentation of the head such as background far behind the subject, very dark or very light background, and appropriately coloured background. 註:某些背景配置可改善頭部影像分割,例如主體後方遠處的背景、非常暗或非常亮的背景,以及顏色適當的背景。
Reflexion on the face caused by the background should not affect the texture rendering of the face skin. 背景造成的臉部反光不應影響臉部皮膚的紋理呈現。
D.3.3.5 Subject and scene lighting D.3.3.5 主體與場景照明
Lighting shall be equally distributed on the skin of the face. There shall be no significant direction of the light from the point of view of the subject. The ratio between the median intensity on a square region centred around landmarks 5.3 and 5.4 with side length 20%20 \% of the inter-eye distance shall be between 0,5 and 2,0. 照明應均勻分佈於臉部皮膚。從主體視角來看,光線不應有明顯的方向性。以地標 5.3 和 5.4 為中心,邊長為眼距 20%20 \% 的正方形區域中,其光線中位強度比應介於 0.5 至 2.0 之間。
Lighting cannot be perfectly uniform and diffuse on the face during acquisition. In order to not create face texture inconstancy by a head movement during the 3D face image acquisition, any head movement should be avoided. 在擷取過程中,臉部照明無法做到完全均勻且漫射。為避免在 3D 臉部影像擷取時,因頭部移動造成臉部紋理不一致,應避免任何頭部移動。
In case of pattern projection during the acquisition, the projected patterns should not be perceivable by the subject in order to not perturbate the subject expression. 若在擷取過程中投射圖案,投射圖案不應被主體察覺,以免干擾主體表情。
For RGB acquisition, the sensor of the camera shall capture the entire visible wavelength, basically the wavelength between 400 nm and 700 nm . This allows correct rendering of the natural colours as seen by humans. Unnaturally coloured lighting, i. e., yellow, red, etc., shall not be used. Care should be taken to adjust the white balance. A high colour rendering index is recommended for illumination. 對於 RGB 影像擷取,相機的感測器應捕捉整個可見光波長,基本上是 400 奈米至 700 奈米之間的波長。這使得影像能正確呈現人類所見的自然色彩。不應使用非自然色彩的照明,例如黃色、紅色等。應注意調整白平衡。建議照明使用高演色性指數。
Illumination shall not cause any red eye effect visible in the eyes and should not cause other lighting artefacts such as spots from a ring flash reducing the visibility of the eyes. 照明不應在眼睛中造成任何紅眼效應,也不應造成其他照明偽影,例如環形閃光燈的光斑,進而降低眼睛的可見度。
The enrolment should be made in a controlled scene; the image should be captured with high signal-tonoise ratio. Noise is not information contained in the original scene but is created by the electronics due to a too high level of amplification. ISO sensitivity settings at values of ISO 100 and ISO 200 typically reduce noise; for high-quality cameras ISO 400 and ISO 800 may also be used. Noise can be minimized by correct exposure at a low ISO setting. 註冊應在受控場景中進行;影像應以高訊號雜訊比擷取。雜訊並非原始場景中包含的資訊,而是由電子設備因過高的放大程度所產生。ISO 感光度設定在 ISO 100 和 ISO 200 的值通常能減少雜訊;對於高品質相機,ISO 400 和 ISO 800 也可能適用。透過在低 ISO 設定下正確曝光,可以將雜訊降至最低。
D.3.3.6 Hot spots and specular reflections D.3.3.6 熱點與鏡面反射
Hot spots (i.e., bright regions that result from light shining directly on the face) shall be absent. These artefacts typically occur when a high intensity focused light source is used for illumination. Diffused lighting, multiple balanced sources or other lighting methods shall be used. 應避免出現熱點(即光線直接照射臉部所造成的明亮區域)。這些瑕疵通常發生在高強度聚焦光源用於照明時。應使用漫射照明、多個平衡光源或其他照明方法。
A single bare point shaped light source like a flash mounted on the 3D acquisition system is not acceptable for imaging. Instead, the illumination should be accomplished using other methods that meet the requirements specified in this clause. 不允許使用單一裸露點狀光源(例如安裝在 3D 採集系統上的閃光燈)進行成像。相反地,應使用符合本條款所規定要求的其他方法來完成照明。
D.3.3.7 Eye glasses D.3.3.7 眼鏡
Enrollees shall not wear eye glasses during acquisition. Enrolment systems shall provide a smart user interface for enrollees who usually wear glasses considering their temporarily reduced reading capabilities. 註冊者在擷取期間不應配戴眼鏡。註冊系統應為通常配戴眼鏡的註冊者提供智慧型使用者介面,考量其暫時降低的閱讀能力。
NOTE Eye glasses interfere with the 3D sensor with respect to several aspects. They hide some parts of the face. The lens might not be completely transparent for the capturing illumination leading to potentially wrong volume determination. The light direction emitted or captured by the 3D system acquisition is perturbed by the lenses. 註:眼鏡會從多個方面干擾 3D 感測器。它們會遮擋臉部某些部分。鏡片可能對擷取照明不完全透明,導致潛在的錯誤體積判斷。3D 系統擷取發射或擷取的光線方向會被鏡片擾亂。
An exception applies when the subject asserts a medical reason to retain the glasses; in these cases, the dark glasses element in the Properties element shall be set to true, even if the glasses are not dark. 當受試者聲稱有醫療原因必須保留眼鏡時,則適用例外情況;在這些情況下,即使眼鏡不是深色的,屬性元素中的深色眼鏡元素也應設定為 true。
D.3.3.8 Shadows in eye-sockets D.3.3.8 眼窩陰影
There should be no shadows in the eye-sockets caused by the eyebrows. The iris and pupil of the eyes should be visible. 眉毛不應在眼窩處造成陰影。眼睛的虹膜和瞳孔應清晰可見。
NOTE This recommendation is intended to exclude images in which the eyes are closed (e.g. during a blink) or half closed. 註:此建議旨在排除眼睛閉合(例如眨眼時)或半閉合的影像。
D.3.3.9 Head coverings D.3.3.9 頭部覆蓋物
In cases where head coverings are present the head coverings element in the Properties block shall be set to true. 若存在頭部覆蓋物,則應將「屬性」區塊中的「頭部覆蓋物」元素設為 true。
Head coverings and shadows caused by head coverings should be absent. An exception applies to cases in which a subject cannot remove a headdress, veil or scarf (e.g., for religious reasons). In such cases the capture process should minimize shadows and obscuration of the face features in the face region. This might involve adjustment of the head coverings. 頭部覆蓋物及其造成的陰影應避免出現。但受測者無法移除頭飾、面紗或圍巾(例如:基於宗教原因)的情況則屬例外。在這些情況下,擷取過程應盡量減少臉部區域的陰影和臉部特徵的遮蔽。這可能需要調整頭部覆蓋物。
D.3.3.10 Visibility of pupils and irises D.3.3.10 瞳孔和虹膜的可見度
In cases where pupils or irises are not visible the pupil or iris not visible element in the Properties block shall be set to true. 若瞳孔或虹膜不可見,則應將「屬性」區塊中的「瞳孔或虹膜不可見」元素設為「真」。
D.3.3.11 Lighting artefacts D.3.3.11 光照偽影
There shall be no lighting artefacts visible on the skin of the face. 臉部皮膚上不得有可見的光照偽影。
D.3.3.12 Eye patches D.3.3.12 眼罩
Eye patches shall not be worn. An exception applies when the subject asserts a need to retain the patch (e.g., for a medical reason); in this case, the left eye patch or the right eye patch element in the Properties block shall be true. 不得佩戴眼罩。當受測者聲明需要保留眼罩時(例如,出於醫療原因),則適用例外情況;在此情況下,屬性區塊中的左眼罩或右眼罩元素應為真。
ISO/IEC 39794-5:2019(E)
D.3.3.13 Hair covering face D.3.3.13 遮臉的頭髮
Hair should not hide parts of the face and should not bring any shadow on the face. 頭髮不應遮蓋臉部,也不應在臉部造成任何陰影。
Hair should not cover any part of the eyes. It is recommended that hair should not cover landmarks 3.2, 3.8, 3.12 for the right eye and landmarks 3.1,3.73.1,3.7 and 3.11 for the left eye, as well as region above these points that measures 5%5 \% of inter-eye distance. 頭髮不應遮蓋眼睛的任何部分。建議頭髮不應遮蓋右眼的標誌點 3.2、3.8、3.12,以及左眼的標誌點 3.1,3.73.1,3.7 和 3.11,以及這些點上方測量為眼距 5%5 \% 的區域。
During acquisition, hair should be swept back behind the ears and above the middle of the forehead whenever possible. 在擷取過程中,應盡可能將頭髮向後梳至耳後並高於額頭中央。
D.3.4 3D acquisition system requirements and recommendations D.3.4 3D 擷取系統要求與建議
D.3.4.1 Purpose D.3.4.1 目的
This clause specifies photographic constraints for the capture of a 3D textured face image. Rather than impose a particular hardware and lighting capture system, this clause specifies the type of output from these systems that is expected. 本條款規定了擷取 3D 紋理臉部影像的攝影限制。本條款並未強制規定特定的硬體和照明擷取系統,而是規定了這些系統預期的輸出類型。
Note that for digital images the normative requirements related to the minimum inter-eye distance impose further requirements on the minimum head size. 請注意,對於數位影像,與最小眼距相關的規範要求對最小頭部尺寸提出了進一步的要求。
D.3.4.2 Contrast and saturation D.3.4.2 對比度和飽和度
For each patch of skin on the capture subject’s face, the gradations in textures shall be clearly visible, i.e., being of reasonable contrast. In this sense, there shall be no saturation (over or under exposure) on the face. 對於擷取對象臉部上的每個皮膚區塊,紋理的漸變應清晰可見,即具有合理的對比度。從這個意義上說,臉部不應出現飽和(過度或曝光不足)現象。
The colour saturation of a 24-bit colour image should be such that after conversion to greyscale, there are 7 bits of intensity variation in the face region of the image. 24 位元彩色影像的色彩飽和度應使其在轉換為灰階後,影像臉部區域具有 7 位元的強度變化。
D.3.4.3 Focus and depth of field D.3.4.3 對焦與景深
The subject’s captured image shall always be in focus from nose to ears and chin to crown. Although this may result in the background behind the subject being out of focus, this is not a problem. 擷取對象的影像應始終從鼻子到耳朵、下巴到頭頂都清晰對焦。儘管這可能導致對象後方的背景失焦,但這並不是問題。
In a typical photographic situation, for optimum quality of the captured face, the f-stop of the lens should be set at two (or more) f-stops below the maximum aperture opening when possible to obtain enough depth of field. 在典型的攝影情境中,為獲得最佳的人臉影像品質,鏡頭的光圈應設定在比最大光圈開口小兩級(或更多)的光圈值,以在可能的情況下獲得足夠的景深。
All images shall have sufficient depth of focus to maintain better than two millimetres spatial sampling rate on the subject’s face features at time of capture. 所有影像應具有足夠的焦深,以在拍攝時保持受試者臉部特徵的空間取樣率優於兩毫米。
The focus and depth of field of the camera shall be set so that the subject’s area scanned is in focus. 相機的對焦和景深應設定為使受試者掃描區域處於對焦狀態。
A depth of field shall be at minimum 150 mm and should be 250 mm or even more. 景深應至少為 150 毫米,且應為 250 毫米或甚至更長。
Greater than one millimetre spatial sampling rate will be considered accomplished if the individual millimetre markings of rulers placed on the subject’s nose and ear facing the camera can be seen simultaneously in a captured test image. 如果將放置在受測者鼻子和耳朵上、面向攝影機的尺規之個別毫米標記,能同時在擷取的測試影像中看到,則視為已達成大於一毫米的空間取樣率。
If the camera lacks auto focus all subject positions will need to be maintained in a defined area for all image captures. 如果攝影機缺乏自動對焦功能,則所有受測者位置在所有影像擷取過程中,都需維持在一個定義的區域內。
D.3.4.4 Greyscale density D.3.4.4 灰階密度
The dynamic range of the image should have at least 7 bits of intensity variation (span a range of at least 128 unique values) in the face region of the image. The face region is defined as the region from crown to chin and from the left ear to the right ear. This recommendation may require 3D acquisition 影像的動態範圍在臉部區域中應至少有 7 位元的強度變化(跨越至少 128 個獨特值的範圍)。臉部區域定義為從頭頂到下巴,以及從左耳到右耳的區域。此建議可能需要 3D 擷取。
system settings to be changed on an individual basis when the skin tone is excessively lighter or darker than the average (present) population. 當膚色比平均(目前)人口過度偏淺或偏深時,可個別更改系統設定。
D.3.4.5 Unnatural colour D.3.4.5 不自然的顏色
Unnaturally coloured lighting, yellow, red, etc., shall not be used. Care shall be taken to correct the white balance of image capture devices. The lighting shall produce a face image with naturally looking flesh tones when viewed in typical examination environments. Images showing the red-eye effect, i.e., the common appearance of red eyes on photographs taken with a photographic flash when the flash is too close to the lens, are not acceptable. The iris and the iris colour shall be visible. 不得使用不自然顏色的照明,例如黃色、紅色等。應注意校正影像擷取裝置的白平衡。照明應在典型的檢查環境中觀看時,產生具有自然膚色的臉部影像。顯示紅眼效應的影像,即當閃光燈離鏡頭太近時,使用攝影閃光燈拍攝的照片上常見的紅眼現象,是不可接受的。虹膜和虹膜顏色應可見。
D.3.4.6 Colour or greyscale enhancement D.3.4.6 顏色或灰階增強
A process that overexposes or under-develops a colour or greyscale image for purposes of beauty enhancement or artistic pleasure is not allowed. The full spectrum shall be represented on the face image where appropriate. Teeth and whites of eyes shall be clearly light or white (when appropriate) and dark hair or features (when appropriate) shall be clearly dark. 不允許為了美觀或藝術目的,對彩色或灰階影像進行過度曝光或顯影不足的處理。臉部影像應適當地呈現完整的光譜。牙齒和眼白應清晰地呈現淺色或白色(如適用),深色頭髮或特徵(如適用)應清晰地呈現深色。
D.3.4.7 Colour calibration D.3.4.7 顏色校準
Colour calibration using an 18%18 \% grey background or other method (such as white balancing) is recommended. 建議使用 18%18 \% 灰色背景或其他方法(例如白平衡)進行色彩校準。
D.3.4.8 Geometrical distortion and CSD of the 3D acquisition system D.3.4.8 3D 擷取系統的幾何失真和 CSD
By definition, a 3D textured acquisition system is designed to precisely measure the shape and the texture of the face. The 3D scanned face data is an accurate representation of the true appearance of the face. 根據定義,3D 紋理擷取系統旨在精確測量臉部的形狀和紋理。3D 掃描的臉部資料是臉部真實外觀的精確呈現。
Geometrical distortion like magnification distortion and radial distortion which appears on conventional 2D acquisition system is compensated by a 3D acquisition system. This compensation is possible because each measured vertex has a known three-dimensional position and viewing angle. 傳統 2D 擷取系統中出現的幾何失真,例如放大失真和徑向失真,可透過 3D 擷取系統進行補償。這種補償之所以可能,是因為每個測量到的頂點都具有已知的三維位置和視角。
For a 3D acquisition system, this Annex doesn’t specify a range for the camera subject distance. The 3D technology of a specific system might have implications on the required camera subject distance in order to achieve the requested accuracy. For many 3D face acquisition systems, this distance is between a few and some dozens of centimetres. For example, passive aero-triangulation 3D acquisition systems used for face acquisition request short distance like 50 cm in order to increase the parallax accuracy and to decrease the general size of the system. 對於 3D 擷取系統,本附錄未指定相機與主體距離的範圍。特定系統的 3D 技術可能會影響所需的相機與主體距離,以達到所需的準確度。對於許多 3D 臉部擷取系統,此距離介於數公分到數十公分之間。例如,用於臉部擷取的被動式空中三角測量 3D 擷取系統需要短距離,例如 50 公分,以提高視差準確度並縮小系統的整體尺寸。
NOTE The 3D textured representation data can be projected to a 2D representation data just by defining the lighting and the virtual camera properties, i. e., the sensor resolution, the photo site size, the focal length, the camera angle, and the camera subject distance. As consequence, the magnification distortion can be simulated from a 3D textured face image representation which contains the true 3D geometrical representation of the face. 註:3D 紋理表示資料可以透過定義照明和虛擬相機屬性(即感測器解析度、感光點尺寸、焦距、相機角度和相機與主體距離)來投影到 2D 表示資料。因此,可以從包含臉部真實 3D 幾何表示的 3D 紋理臉部影像表示中模擬放大失真。
D.3.4.9 Focal length D.3.4.9 焦距
The selection of a camera and its lens is a major factor affecting the quality of face images. To ensure high image quality and a standard compliant inter-eye distance (IED), the camera’s sensor must have sufficient pixel dimensions and its lens must be chosen to match its image sensor’s physical dimensions. 攝影機及其鏡頭的選擇是影響臉部影像品質的主要因素。為確保高影像品質和符合標準的眼距(IED),攝影機的感測器必須具有足夠的像素尺寸,且其鏡頭必須選擇與其影像感測器的實體尺寸相符。
For a selected CSD (in millimetres), and acamera image sensor with a vertical dimension in millimetres of h_(mm)h_{m m} a requested vertical field of view of H_("FieldofView ")H_{\text {FieldofView }} (in millimetres) and the focal length ff (in millimetres) can be computed using the following relationship in order to optimise the requested field of view of the subject with respect to the sensor dimensions: 對於選定的 CSD(以毫米計),以及垂直尺寸為 h_(mm)h_{m m} 毫米的攝影機影像感測器,所要求的垂直視野為 H_("FieldofView ")H_{\text {FieldofView }} (以毫米計),焦距 ff (以毫米計)可使用以下關係式計算,以最佳化受測者相對於感測器尺寸的所需視野:
f~=h_(mm)(CSD)/(H_("FieldOfView "))f \cong h_{m m} \frac{C S D}{H_{\text {FieldOfView }}}
ISO/IEC 39794-5:2019(E)
D.3.4.10 Sensor resolution in pixel D.3.4.10 感測器像素解析度
The resolution of the textured images acquired by a 2 D CMOS/CCD sensor can be computed in the following way: For a camera image sensor with vertical pixel count of h_(px)h_{p x}, the inter-eye distance on the sensor in pixels IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} may be computed using the following relationship, where IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} is the intereye distance in millimetres on the subject: 透過 2D CMOS/CCD 感測器擷取之紋理影像的解析度可透過以下方式計算:對於垂直像素計數為 h_(px)h_{p x} 的相機影像感測器,感測器上像素間的眼距 IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} 可使用以下關係式計算,其中 IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} 為受測物上以毫米計的眼距:
IED_(mm)^("Sensor ")=IED_(mm)^("Subject ")(f)/(CSD)I E D_{m m}^{\text {Sensor }}=I E D_{m m}^{\text {Subject }} \frac{f}{C S D}
and 和
IED_(px)^("Sensor ")=IED_(mm)^("Sensor ")(h_(px))/(h_(mm))I E D_{p x}^{\text {Sensor }}=I E D_{m m}^{\text {Sensor }} \frac{h_{p x}}{h_{m m}}
EXAMPLE A camera has the following specification: APS-C sensor, 22,3mmxx14,9mm,2592pxxx1944px22,3 \mathrm{~mm} \times 14,9 \mathrm{~mm}, 2592 \mathrm{px} \times 1944 \mathrm{px}, 5 megapixels. For a CSD of 600 mm , a typical H_("FieldOfView ")H_{\text {FieldOfView }} of 500 mm , a typical IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} of about 62 mm the calculations below show that the focal length ff will be about 25 mm . Then IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} will be about 336 pixels, well above the requirement of 240 pixels. The finest detail measured by each pixel of the camera is (H_("FieldOfView "))/(h_(px))\frac{H_{\text {FieldOfView }}}{h_{p x}} which corresponds for this example to around 0,2mm0,2 \mathrm{~mm}. 範例 相機具有以下規格:APS-C 感測器、 22,3mmxx14,9mm,2592pxxx1944px22,3 \mathrm{~mm} \times 14,9 \mathrm{~mm}, 2592 \mathrm{px} \times 1944 \mathrm{px} 、5 百萬像素。對於 600 毫米的 CSD、500 毫米的典型 H_("FieldOfView ")H_{\text {FieldOfView }} 、約 62 毫米的典型 IED_(mm)^("Subject ")I E D_{m m}^{\text {Subject }} ,以下計算顯示焦距 ff 約為 25 毫米。則 IED_(px)^("Sensor ")I E D_{p x}^{\text {Sensor }} 約為 336 像素,遠高於 240 像素的要求。相機每個像素測得的最精細細節為 (H_("FieldOfView "))/(h_(px))\frac{H_{\text {FieldOfView }}}{h_{p x}} ,在此範例中約為 0,2mm0,2 \mathrm{~mm} 。
D.3.5 Digital requirements and recommendations D.3.5 數位要求與建議
D.3.5.1 Geometry D.3.5.1 幾何
D.3.5.1.1 Texture map geometry D.3.5.1.1 紋理貼圖幾何
The origin of coordinates of the texture map shall be at the upper left given by coordinate (0,0)(0,0) with positive entries from left to right (first dimension) and top to bottom (second dimension). 紋理貼圖的座標原點應位於左上角,由座標 (0,0)(0,0) 給定,其正值從左到右(第一維)和從上到下(第二維)遞增。
D.3.5.1.2 3D data geometry D.3.5.1.2 3D 資料幾何
The vertex origin after scale and offset application shall be the centre of the two eyes, i.e., the midpoint between the left eye centre (12.1) and the right eye centre (12.2). 套用縮放和偏移後的頂點原點應為雙眼中心,即左眼中心 (12.1) 和右眼中心 (12.2) 之間的中點。
The measuring unit used for 3D represention face image data after scale application shall be millimetre. 用於縮放應用後 3D 人臉影像資料的測量單位應為公釐。
The face orientation after scale and offset application shall follow: 縮放和偏移應用後的人臉方向應遵循:
Horizontal axis (x), passing between the left eye centre (12.1) and the right eye centre (12.2), and oriented to the left eye centre direction. 水平軸(x),通過左眼中心(12.1)和右眼中心(12.2)之間,並朝向左眼中心方向。
Depth axis (z), defined by the standard position of the face which is determined when the head is in the rest position and the eye is looking straight forward. 深度軸(z),由人臉的標準位置定義,該位置是在頭部處於靜止位置且眼睛直視前方時確定的。
Vertical axis (y), defined by the right hand rule from the two other axis. 垂直軸(y),由其他兩個軸依循右手定則定義。
The pitch of the face in rest position corresponds to common face acquisition. This pitch is slightly different to the pitch defined by the frontal pose which is associated to the Frankfurt Horizon. The Frankfurt Horizon rule is not adapted here for several reasons. The head pitch of the Frankfurt Horizon doesn’t correspond to general attitude in rest position of all ethnic groups and involves discomfort and bad eye position in the orbit (Figure D.39). The Frankfurt Horizon is related to the ear position which might not be not scanable as it is, e.g., covered by hair. 臉部在靜止位置的俯仰角與常見的臉部擷取方式相符。此俯仰角與由額頭姿勢定義的俯仰角略有不同,後者與法蘭克福水平線相關。法蘭克福水平線規則在此不適用,原因有幾點。法蘭克福水平線的頭部俯仰角與所有族群在靜止位置的普遍姿態不符,並會導致不適和眼眶中眼睛位置不佳(圖 D.39)。法蘭克福水平線與耳朵位置相關,而耳朵位置可能無法掃描,例如被頭髮覆蓋。
a) Head orientation to the Frankfurt Horizon doesn't correspond to the eye direction a) 頭部朝向法蘭克福水平線的方向與眼睛方向不符
b) Head pitch is oriented according to the rest position: the face pose doesn’t follow the Frankfurt Horizon b) 頭部俯仰角依靜止位置調整:臉部姿勢不遵循法蘭克福水平線
Key 圖例
1 Frankfurt Horizon 1 法蘭克福水平線
2 eye direction 2 眼睛方向
3 lowest point of the right eye socket 3 右眼窩最低點
4 tragion 4 耳屏上切跡
Figure D. 39 - Frankfurt Horizon in relation to eye position/direction 圖 D. 39 - 法蘭克福水平線與眼睛位置/方向的關係
D.3.5.2 Colour profile D.3.5.2 色彩設定檔
3D textured images shall be represented as one of the following. The captured image shall be a truecolour representation of the subject in a typical colour space such as sRGB as specified in IEC 61966-2. Other true-colour representations may be used, but in all cases the ICC colour profile shall be embedded inside the textured map for all formats (JPEG, JPEG-2000 and PNG): 3D 紋理影像應以下列其中一種方式呈現。擷取的影像應為受測物在典型色彩空間(例如 IEC 61966-2 中指定的 sRGB)中的真彩色表示。可以使用其他真彩色表示,但在所有情況下,ICC 色彩設定檔都應嵌入所有格式(JPEG、JPEG-2000 和 PNG)的紋理圖中:
a) 24 bit or 48 -bit RGB colour space where for every pixel, 8 bits or 16 bits will be used to represent each of the red, green, and blue components. a) 24 位元或 48 位元 RGB 色彩空間,其中每個像素的紅色、綠色和藍色元件將分別使用 8 位元或 16 位元來表示。
b) 8 bit or 16 -bit monochrome colour space where for every pixel, 8 bits or 16 bits will be used to represent the luminance component. b) 8 位元或 16 位元單色色彩空間,其中每個像素的亮度元件將使用 8 位元或 16 位元來表示。
RGB acquisition is recommended. 建議採用 RGB 擷取。
Colour quality should be measured in terms of colour error using the CIEDE2000 formula (deltaE2000) of a standardized test pattern. The average deltaE2000 of all colour patches should not exceed 10 for camera systems. The maximum deltaE2000 for any colour patch should not exceed 20 for camera systems. Measured CIELAB Lab* human skin tone a^(**)a^{*} and b^(**)b^{*} values shall be positive. Negative a^(**)a^{*} and b^(**)b^{*} values are acceptable only for medical reasons. 色彩品質應使用標準化測試圖案的 CIEDE2000 公式 (deltaE2000) 以色彩誤差來衡量。攝影機系統的所有色塊平均 deltaE2000 不應超過 10。攝影機系統的任何色塊最大 deltaE2000 不應超過 20。測量的 CIELAB Lab* 人類膚色 a^(**)a^{*} 和 b^(**)b^{*} 值應為正值。負 a^(**)a^{*} 和 b^(**)b^{*} 值僅在醫療原因下可接受。
Interlaced video frames shall not be used for the 3D textured image type. All interlacing shall be absent. 3D 紋理影像類型不得使用交錯視訊影格。所有交錯都應不存在。
D.3.5.3 Use of near infrared cameras D.3.5.3 近紅外線攝影機之使用
If dedicated near infrared cameras are used, one should be aware that the interoperability between white light and near infrared images might be reduced. 如果使用專用的近紅外線攝影機,應注意白光影像和近紅外線影像之間的互通性可能會降低。
ISO/IEC 39794-5:2019(E)
D.3.5.4 Spatial sampling rate D.3.5.4 空間取樣率
When acquisition is done in visible spectra the spatial sampling rate of the texture map shall be such that the IED is at least 240 pixels. 當在可見光譜中進行採集時,紋理圖的空間取樣率應使 IED 至少為 240 像素。
The 3D representation data shall be able to measure shape variation of a size of less than 5 mm on all axes and should be able to measure shape variation of a size of less than 2 mm on all axes. 3D 表示資料應能測量所有軸上小於 5 毫米的形狀變化,並應能測量所有軸上小於 2 毫米的形狀變化。
D.3.5.5 Post-acquisition processing D.3.5.5 後期處理
No post-processing other than creation of the 3D representation data and the corresponding texture, 3D rotation, cropping, downsampling and/or multiple compressions shall be applied to derive a 3D face image from a captured image. Multiple (i.e., repeated) compressions should be avoided when generating 3D textured map images. 除了建立 3D 表示資料和對應的紋理、3D 旋轉、裁切、降取樣和/或多次壓縮之外,不得對擷取的影像進行任何後處理,以產生 3D 臉部影像。在產生 3D 紋理貼圖影像時,應避免多次(即重複)壓縮。
The calibration accuracy of the acquisition device shall be high enough such that the mean shift between the texture of the 2D image and the 3D data is less than 1 mm . 採集設備的校準準確度應夠高,以使二維影像的紋理與三維資料之間的平均位移小於 1 毫米。
D.3.6 Requirements on 3D textured image representation D.3.6 3D 紋理影像表示法的要求
D.3.6.1 Coordinate system type D.3.6.1 座標系統類型
The coordinate system type shall be Cartesian. Vertex coordinates are positive and coded with no decimal inside a range from 0 to 65535 (unsigned short). 座標系統類型應為笛卡兒座標。頂點座標為正值,編碼時不帶小數點,範圍介於 0 到 65535 之間(無符號短整數)。
D.3.6.2 Scales and offsets D.3.6.2 比例與偏移
The face centre shall be at origin after application of scales factor and the offset. As consequence, the 3 Offsets OffsetX, OffsetY and OffsetZ shall be negative. 臉部中心在套用比例因子和偏移後應位於原點。因此,三個偏移量 OffsetX、OffsetY 和 OffsetZ 應為負值。
The transformation to metric coordinates is described by appropriate scaling factors. The unit of the scale factor is the millimetre. The scaling factors shall be the same on all three axes. (ScaleX = ScaleY = ScaleZ). The vertex precision in millimetres is given by the scale value. 轉換為公制座標是透過適當的比例因子來描述。比例因子的單位是毫米。三個軸上的比例因子應相同。(ScaleX = ScaleY = ScaleZ)。以毫米為單位的頂點精度由比例值給出。
As an example, if the scaling factor is set to 0,1mm0,1 \mathrm{~mm}, then the vertex encoding leads to a precision of 0,1mm0,1 \mathrm{~mm}, and the range shape covers a cube of around 2768mm+-3mm2768 \mathrm{~mm} \pm 3 \mathrm{~mm} which allows all face and body encoding. This scaling factor of 0,1mm0,1 \mathrm{~mm} should be of sufficient precision for most biometric applications. 舉例來說,如果比例因子設定為 0,1mm0,1 \mathrm{~mm} ,則頂點編碼會產生 0,1mm0,1 \mathrm{~mm} 的精度,且範圍形狀涵蓋約 2768mm+-3mm2768 \mathrm{~mm} \pm 3 \mathrm{~mm} 的立方體,這允許所有面部和身體編碼。這個 0,1mm0,1 \mathrm{~mm} 的比例因子對於大多數生物識別應用來說,應具有足夠的精度。
D.3.6.3 Vertex information D.3.6.3 頂點資訊
The vertex identifier shall not be present. The vertex identifier is implicitly given by the stack order of the vertex list. The first vertex shall have the index 0 . 不應存在頂點識別碼。頂點識別碼由頂點列表的堆疊順序隱含給定。第一個頂點的索引應為 0。
The index defined in the vertex triangle data block refers to this implicit order. 頂點三角形資料區塊中定義的索引指的是這個隱含的順序。
Three vertices, connected to each other by three edges, define a face called triangle. These vertices are listed in counter clockwise order when looking at the face from the outside. This rule defines the normal of each triangle. Each triangle shall share two vertices with each of its adjacent triangles. 三個頂點,透過三條邊相互連接,定義了一個稱為三角形的面。從外部觀察該面時,這些頂點以逆時針順序排列。此規則定義了每個三角形的法線。每個三角形應與其相鄰的每個三角形共享兩個頂點。
If there is any hole on the pupils, brows or nostrils, it may be filled. For enrolment, the 3D shape composed by all the triangles shall not have any hole inside the area of the front of the head. The usage of the Vertex textures block is mandatory. 如果瞳孔、眉毛或鼻孔有任何孔洞,可以將其填補。對於註冊,由所有三角形組成的 3D 形狀在頭部前方區域內不得有任何孔洞。頂點紋理區塊的使用是強制性的。
The vertex normals block shall not be present. If necessary, vertex normals can be recomputed from the normals of its neighbours. 頂點法線區塊不得存在。如有必要,頂點法線可從其鄰近的法線重新計算。
D.3.6.4 Texture map D.3.6.4 紋理貼圖
For biometric recognition purposes, either shape data without texture data, or a combination of shape data and texture data can be used. As a consequence, the texture map is optional. 為生物辨識目的,可使用不含紋理資料的形狀資料,或形狀資料與紋理資料的組合。因此,紋理貼圖為選用。
The Texture projection matrix block shall not be used as each vertex texture value is already defined by the spatial coordinates UV which refer to the texture map. 紋理投影矩陣區塊不應使用,因為每個頂點紋理值已由參考紋理貼圖的空間座標 UV 定義。
The acquisition period is defined by the absolute difference between the acquisition end and the acquisition start time. During this period neither the subject nor the acquisition system shall move or be moved. 採集期間由採集結束時間與採集開始時間的絕對差值定義。在此期間,受測者和採集系統均不得移動或被移動。
D.3.6.5 Face area scanned D.3.6.5 臉部掃描區域
The Face area scanned element shall include the area of the head which is covered by the scanned data. It may have the following values: front of the face, ears, chin, neck, back of the face or full head. The presence of the front of the face is mandatory. 臉部掃描區域元素應包含掃描資料所涵蓋的頭部區域。其可能的值為:臉部正面、耳朵、下巴、頸部、臉部背面或整個頭部。臉部正面的存在是強制性的。
Annex E
(informative) 附錄 E (參考性)
Additional technical considerations 額外的技術考量
E. 1 Setup examples for face portrait capturing E.1 臉部肖像擷取設定範例
E.1.1 General E.1.1 一般規定
The implementation of the face portrait acquisition setup should be done considering the properties of the different technologies and environments. Some face portrait acquisition setups are detailed below. This annex lists examples in no particular order and does neither recommend technologies mentioned here nor exclude technologies not mentioned here. 臉部肖像擷取設定的實施應考量不同技術和環境的特性。以下詳述了一些臉部肖像擷取設定。本附錄所列範例不分先後順序,既不推薦此處提及的技術,也不排除此處未提及的技術。
E.1.2 Studio environment with one single light E.1.2 攝影棚環境與單一光源
A single light and multiple reflector panels illuminate the subject’s face uniformly. The light with a reflector should be placed approximately 35^(@)35^{\circ} above the line between the camera and the subject and be directed toward the subject’s face at a horizontal angle of less than 45^(@)45^{\circ} from the line. A reflector panel should be placed on the subject’s opposite side to prevent shadows on the face. Optionally, an additional reflector may be placed below and in front of the subject’s face to illuminate the area around the chin. See Figure E.1. 單一光源和多個反光板均勻地照亮受測者的臉部。帶有反光板的光源應放置在攝影機與受測者之間連線上方約 35^(@)35^{\circ} 處,並以小於該連線 45^(@)45^{\circ} 的水平角度指向受測者的臉部。反光板應放置在受測者臉部的另一側,以防止產生陰影。此外,可選擇在受測者臉部下方和前方放置額外的反光板,以照亮下巴周圍的區域。請參閱圖 E.1。
Figure E. 1 - Example of a setup with one single light 圖 E.1 – 單一光源設置範例
E.1.3 Studio environment with two lights E.1.3 雙光源攝影棚環境
Two lights with reflectors should be placed approximately 35^(@)35^{\circ} above the line between the camera lens and the subject. Both lights should be placed within 45^(@)45^{\circ} of the line between the camera lens and the subject. The optional plane reflector in front of the subject supplies additional light around and below the subject’s chin. See Figure E.2. 兩個帶有反光板的光源應放置在攝影機鏡頭與受測者之間連線上方約 35^(@)35^{\circ} 處。兩個光源都應放置在攝影機鏡頭與受測者之間連線的 45^(@)45^{\circ} 範圍內。受測者前方可選用的平面反光板可為受測者下巴周圍和下方提供額外光源。請參閱圖 E.2。
Figure E. 2 - Example of a setup with two front lights 圖 E.2 – 雙前燈設置範例
E.1.4 Studio environment with two lights and background illumination E.1.4 雙燈及背景照明攝影棚環境
A background light is added to the arrangement in E.1.3 to eliminate shadows visible on the background behind the head. The background light should be targeted to the background and be placed directly behind and below the subject. See Figure E.3. 在 E.1.3 的配置中增加背景燈,以消除頭部後方背景上可見的陰影。背景燈應對準背景,並直接放置在受測者後方和下方。請參閱圖 E.3。
Figure E. 3 - Example of a setup with two front lights and one background light 圖 E.3 – 雙前燈及單一背景燈設置範例
E.1.5 Photo booth environment E.1.5 照相亭環境
The requirements for a good face portrait also apply for a photo booth and an operator should make their best effort to get as close as possible to the recommendations given in this document. In studio environments, the human element is capable of checking the quality assurance; this capability should be replaced by automated quality assurance technology in photo booths and kiosks. 良好的臉部肖像要求也適用於照相亭,操作員應盡最大努力,盡可能接近本文件中的建議。在攝影棚環境中,人為因素能夠檢查品質保證;此能力應由照相亭和自助服務機中的自動化品質保證技術取代。
In a photo booth, multiple lights should be positioned symmetrically behind a diffuser panel above and aside of the camera to provide uniform lighting on the subject’s face and to eliminate glare and shadows visible on the face. See Figures E. 4 and E.7. A background light should be placed on the ground between the background and the subject. The front lights should be placed at an angle of approximately 35^(@)35^{\circ} above the line between the camera and the subject’s head to prevent reflection artefacts on the 在照相亭中,多個燈光應對稱地放置在相機上方和旁邊的擴散板後面,以在受試者臉部提供均勻照明,並消除臉部可見的眩光和陰影。請參見圖 E.4 和 E.7。背景燈應放置在背景和受試者之間的地面上。前燈應以約 35^(@)35^{\circ} 的角度放置在相機和受試者頭部之間的線上,以防止受試者眼鏡上的反射偽影。
subject’s glasses. The inside walls should be white and serve as reflectors. Directly behind the subject should be no directly reflecting material. The interior lights of the booth should be kept switched on during operation to reduce red eye effects. Direct or indirect lighting from below and in front of the subject should be used to eliminate shadows around the chin. An opaque curtain should be used and be closed during capturing to eliminate external light effects. 受試者的眼鏡。內部牆壁應為白色並作為反射器。受試者正後方不應有直接反射材料。照相亭的內部燈光在操作期間應保持開啟,以減少紅眼效應。應使用來自受試者下方和前方的直接或間接照明,以消除下巴周圍的陰影。應使用不透明的窗簾,並在拍攝期間關閉,以消除外部光線影響。
Proper positioning of the subject and control of the subject’s pose may be improved through feedback provided to the subject via a mirror or a live-video monitor. A height-adjustable seat or camera should be provided to allow the subject to face the camera. See Figure E.5. Alternatively, the camera may be movable to adjust the height to the head position. See Figure E.6. 透過鏡子或即時視訊螢幕向受測者提供回饋,可改善受測者的正確定位及姿勢控制。應提供可調整高度的座椅或攝影機,以使受測者面向攝影機。請參閱圖 E.5。或者,攝影機可移動以調整高度至頭部位置。請參閱圖 E.6。
Figure E.4- Example of a photo booth setup: Front view 圖 E.4 - 照相亭設置範例:前視圖
Figure E. 5 - Example of a photo booth setup: Side view 圖 E.5 - 照相亭設置範例:側視圖
Figure E. 6 - Example of a photo booth setup: Side view with height adjustable camera 圖 E.6 - 照相亭設置範例:具備高度可調攝影機的側視圖
Figure E. 7 - Example of a photo booth setup: Top view 圖 E.7 — 照相亭設置範例:俯視圖
The requirements for a good face portrait also apply in a registration office environment and an operator should try to get as close as possible to the recommendations given in this document. One should have in mind that such an easier setup regularly leads to face portraits of suboptimal quality and should therefore not be the preferred solution. 良好臉部肖像的要求也適用於戶政事務所環境,操作員應盡可能接近本文件中的建議。應記住,這種較為簡便的設置通常會導致品質不佳的臉部肖像,因此不應作為首選解決方案。
The subject and the background should be illuminated by two diffuse light sources that are mounted in a console with a small footprint, so that it fits into a typical registration office environment. The console may be mounted on the floor or on the wall. Flash light should not be used, at most in combination with appropriate permanent illumination. The main illumination during capturing should be that of the capturing system. Illumination mainly by the room illumination from the roof lights, the window or desk light is not acceptable, as well as direct sun light. Even if the office conditions might require much easier setups, the principle requirement of a uniform illumination remains valid. 受試者和背景應由兩個漫射光源照明,這些光源安裝在佔用空間小的控制台中,以便其適合典型的戶政事務所環境。控制台可以安裝在地板上或牆壁上。不應使用閃光燈,最多只能與適當的永久照明結合使用。拍攝期間的主要照明應來自拍攝系統。主要由屋頂燈、窗戶或書桌燈的室內照明,以及直射陽光,都是不可接受的。即使辦公室條件可能需要更簡便的設置,均勻照明的基本要求仍然有效。
A revolving and height-adjustable chair or stool with an additional cushion for smaller capture subjects should be provided to allow the subject to face the camera and adjust his head to the proper height. See Figure E.8. 應提供可旋轉、高度可調的座椅或凳子,並為較小的擷取對象提供額外的坐墊,以使對象能夠面向攝影機並將頭部調整到適當的高度。請參閱圖 E.8。
Figure E. 8 - Example of a registration office environment setup 圖 E. 8 - 註冊辦公室環境設置範例
Feedback should be provided to the subject via a second live-video monitor facing to the subject for positioning and behaviour guidance. An image preview should be offered to allow a subject to choose from a selection of face portraits or to be recaptured if necessary, before the final face portrait is submitted for further processing. 應透過面向對象的第二個即時視訊螢幕向對象提供回饋,以進行定位和行為引導。應提供影像預覽,以便對象可以從一系列臉部肖像中進行選擇,或在必要時重新擷取,然後再提交最終的臉部肖像進行進一步處理。
However, empirical data from production environments indicates that subjects who see the live view enter into a “vanity mode”. This can significantly reduce the throughput of the process and the quality of the captured biometric data. As an alternative to live view, visual, graphical or verbal instructions should be provided to the subject to reach optimal face and body posture. 然而,來自生產環境的實證資料顯示,看到即時影像的對象會進入「虛榮模式」。這會顯著降低處理的吞吐量和擷取到的生物識別資料的品質。作為即時影像的替代方案,應向對象提供視覺、圖形或口頭指示,以達到最佳的臉部和身體姿勢。
E.1.7 Setup with flash E.1.7 閃光燈設置
If carefully applied, flashes may be used. In this case the quality requirements especially with respect to shadows, homogenous illumination, and absence of reflections in the eyes have to be maintained. See Figure E.9. The given distance measures are examples. 如果小心使用,可以使用閃光燈。在這種情況下,必須保持品質要求,特別是在陰影、均勻照明以及眼睛中無反射方面。請參閱圖 E.9。所給的距離測量值為範例。
Figure E. 9 - Example of a setup with flash, top and side view 圖 E.9 - 使用閃光燈的設置範例,俯視圖和側視圖
The given measures are examples. 所給的測量值為範例。
E. 2 Measuring magnification and radial distortion in a face portrait capture setup E.2 測量人臉肖像擷取設定中的放大率和徑向失真
E.2.1 General E.2.1 概述
This annex deals with two kinds of distortion. 本附錄處理兩種失真。
The first kind is the magnification distortion, a geometrical effect of the optical perspective. Some optical systems image objects of the same size differently depending on the distance between object and sensor. Such magnification distortion always appears in human vision. 第一種是放大失真,這是一種光學透視的幾何效應。某些光學系統會根據物體與感測器之間的距離,以不同的方式成像相同大小的物體。這種放大失真總是出現在人類視覺中。
The second kind is the radial distortion (barrel distortion, pincushion distortion, moustache distortion) caused by optical properties of a lens. This feature doesn’t exist for human vision. 第二種是鏡頭光學特性造成的徑向失真(桶狀失真、枕狀失真、鬍鬚狀失真)。人類視覺不存在此特徵。
Both magnification and radial distortion may influence the performance of automated face recognition systems as well as of human recognition. Therefore, this Annex describes how to measure various types of distortion using two targets. 放大率和徑向失真都可能影響自動人臉辨識系統以及人類辨識的效能。因此,本附錄說明如何使用兩個目標來測量各種失真類型。
E.2.2 Magnification distortion target construction E.2.2 放大失真目標建置
To build the magnification distortion target, take Figure E.10, print it on A4 size paper, and fold it into a T shape. One should look at the foot of the T , the head bar is on the opposite side away from the viewer. The length of the T leg is the typical eye to nose distance of a Caucasian ( 50 mm ). In a photograph of the magnification distortion target taken as described in E.2.4, the relative size of an object near to the eye level and near to the nose tip level of a human can be measured. Additionally, it can be observed if the image is sharp enough at the entire face region including nose and eyes looking at the visibility of the 0,5mm0,5 \mathrm{~mm} and 1 mm wide markers in the squares. See Figure E.11. 若要建置放大失真目標,請取圖 E.10,將其列印在 A4 大小的紙張上,然後將其摺疊成 T 形。應查看 T 的底部,橫桿位於觀察者對面的另一側。T 腿的長度是高加索人典型的眼睛到鼻子距離(50 毫米)。在如 E.2.4 所述拍攝的放大失真目標照片中,可以測量靠近眼睛水平和靠近人類鼻尖水平的物件的相對大小。此外,還可以觀察影像在整個臉部區域(包括鼻子和眼睛)是否足夠清晰,方法是查看方塊中 0 和 1 毫米寬標記的可見度。請參閱圖 E.11。
Figure E. 11 - Ready to use simple magnification distortion target 圖 E. 11 - 可立即使用的簡易放大失真目標
In Figure E.11, the millimetre markers on the nose level, the markers in 20 mm distance on nose and eye level, as well as the resolution targets with one and two line pairs per millimetre are shown. 在圖 E.11 中,顯示了鼻部高度的毫米標記、鼻部和眼部高度相距 20 毫米的標記,以及每毫米一對和兩對線條的解析度目標。
To precisely and repeatedly measure magnification distortion, fabrication of a rigid magnification distortion target is recommended. 為了精確且重複地測量放大失真,建議製作一個堅固的放大失真目標。
To keep the magnification distortion target pointing at the camera it can be necessary to attach the folded target to a support. For target support material it is recommended to use white foamboard with a thickness of 5 mm or 2 mm aluminium. The target support structure construction is also suitable for 3D printing using plastic materials in light colour. Paper based materials should not be used. 為了使放大失真目標始終指向攝影機,可能需要將摺疊的目標固定在一個支援物上。對於目標支援材料,建議使用厚度為 5 毫米的白色泡棉板或 2 毫米的鋁材。目標支援結構的建構也適用於使用淺色塑膠材料進行 3D 列印。不應使用紙質材料。
This magnification distortion target should be used to measure magnification distortion in the nose region of the face. The magnification distortion measuring is described in E.2.5. 此放大失真目標應用於測量臉部鼻部區域的放大失真。放大失真測量請參閱 E.2.5。
E.2.3 Radial distortion target construction E.2.3 徑向畸變目標建構
A radial distortion target is composed of evenly spaced, horizontal and vertical lines, forming a net-like pattern. 徑向畸變測試圖由等距的水平線和垂直線組成,形成網狀圖案。
Start building the radial distortion target by cutting the back plane and supporting panels. 開始製作徑向失真目標,首先切割背板和支撐板。
This can be done by stacking four foamboard pieces together to form a 50mmxx160mmxx20mm50 \mathrm{~mm} \times 160 \mathrm{~mm} \times 20 \mathrm{~mm} support slab. See Figure E.12. If thinner aluminium or thicker foamboard is used, then change the dimensions accordingly. Paper based materials should not be used for the support. Foamboard or similar material is recommended for the box material. The size of the visible white board is 200mmxx200mm200 \mathrm{~mm} \times 200 \mathrm{~mm}. To maintain the base material flatness requirements, use 2 mm aluminium or 5 mm foamboard. 這可以透過將四塊泡棉板堆疊在一起,形成一個 50mmxx160mmxx20mm50 \mathrm{~mm} \times 160 \mathrm{~mm} \times 20 \mathrm{~mm} 支撐板來完成。請參見圖 E.12。如果使用較薄的鋁板或較厚的泡棉板,則應相應地更改尺寸。不應使用紙質材料作為支撐。建議使用泡棉板或類似材料作為盒子的材料。可見白板的尺寸為 200mmxx200mm200 \mathrm{~mm} \times 200 \mathrm{~mm} 。為符合基材平整度要求,請使用 2 毫米鋁板或 5 毫米泡棉板。
ISO/IEC 39794-5:2019(E)
Glue the four foamboard pieces together. Use strong glue that does not melt board material. Make sure that the slab has 90^(@)90^{\circ} corners. In order to keep the slab in correct shape use supports while gluing the slab together piece by piece. 將四塊泡棉板黏合在一起。請使用不會融化板材的強力膠。請確保板材有 90^(@)90^{\circ} 個角。為了使板材保持正確形狀,請在逐片黏合板材時使用支撐物。
Figure E. 12 - Magnification distortion target support parts and dimensions for 5 mm thick foamboard construction 圖 E. 12 - 5 毫米厚泡棉板結構的放大失真目標支援部件和尺寸
The printed radial distortion target is glued on the back pane foamboard. Figure E. 13 contains the target support board back plane radial distortion target print version for A4 size printing. 印刷的徑向失真目標黏貼在背板泡棉板上。圖 E. 13 包含目標支援板背板徑向失真目標的 A4 尺寸列印版本。
Print the target on an A4 paper. Check the size of the grid before cutting. The grid size is 150mmxx150mm150 \mathrm{~mm} \times 150 \mathrm{~mm} (exact size is 151,1mmxx151,1mm151,1 \mathrm{~mm} \times 151,1 \mathrm{~mm} due to the line width in use). Cut out the radial distortion target along the outmost border line so that the border line stays intact or cut out the upper and lower part of the A4 to form a 200mmxx200mm200 \mathrm{~mm} \times 200 \mathrm{~mm} size paper target. Glue the radial distortion target at the centre of the 210mmxx210mm210 \mathrm{~mm} \times 210 \mathrm{~mm} foamboard. 將目標列印在 A4 紙上。在裁切前檢查網格尺寸。網格尺寸為 150mmxx150mm150 \mathrm{~mm} \times 150 \mathrm{~mm} (由於使用的線寬,精確尺寸為 151,1mmxx151,1mm151,1 \mathrm{~mm} \times 151,1 \mathrm{~mm} )。沿著最外層邊界線裁切徑向失真目標,使邊界線保持完整,或者裁切 A4 紙的上下部分以形成 200mmxx200mm200 \mathrm{~mm} \times 200 \mathrm{~mm} 尺寸的紙質目標。將徑向失真目標黏貼在 210mmxx210mm210 \mathrm{~mm} \times 210 \mathrm{~mm} 泡棉板的中心。
Figure E. 13 - Radial distortion target version for A4 size printing 圖 E. 13 - 用於 A4 尺寸列印的徑向失真目標版本
The radial distortion target should be used to measure barrel or pincushion distortion. The radial distortion measuring is described in E.2.6. 徑向失真目標應用於測量桶形或枕形失真。徑向失真測量在 E.2.6 中描述。
E.2.4 Target photography E.2.4 目標攝影
Place the target at the typical location of the subject’s head in the given photographical setup or glue it in a target support board setup. Take a picture of the target in the intended setup, from the chosen distance, with the intended focal length, with the intended aperture. 將目標放置在指定攝影設置中受測者頭部的典型位置,或將其黏貼在目標支撐板設置中。在預期設置中,從選定的距離、使用預期的焦距和預期的光圈,拍攝目標的照片。
Target is placed at an appropriate distance from the lens following the guidelines set in this document. Before taking photographs, camera and lights are set following the recommendations of this document. 目標依據本文件所設定的準則,放置在距離鏡頭適當的位置。在拍攝照片之前,相機和燈光會依據本文件的建議進行設定。
It is easy to use a tripod or similar support to keep the target properly aligned. Touch (hook and loop) fasteners may be used to temporarily hold the target for photography. A small patch of fastener tape attached to the target support is able to hold lightweight targets as described in this document. 使用三腳架或類似的支援物可以輕鬆地保持目標正確對齊。魔鬼氈(鉤環)扣件可用於暫時固定目標以進行攝影。一小塊貼在目標支撐物上的魔鬼氈膠帶就能固定本文件所述的輕量目標。
If the distance between the foremost point of a subject and the optical centre of a standard lens (i.e not telecentric) is DD and the height of a structure S_(1)S_{1} in the front of the face, e.g., the nose is h_(s1)\mathrm{h}_{\mathrm{s} 1}, then the camera to subject distance (CSD) is assumed to be Delta D+D\Delta D+D and a structure S_(2)S_{2} of the height Deltah_(s1)+h_(s1)\Delta h_{s 1}+h_{s 1} at eye level, would virtually appear to have the same size as S_(1)*S_(1)S_{1} \cdot S_{1} virtually seems to be larger than it is in reality due to the magnification distortion. See Figure E.14. 如果受測者最前端點與標準鏡頭(即非遠心鏡頭)光學中心之間的距離為 DD ,且臉部前方結構(例如鼻子)的高度為 h_(s1)\mathrm{h}_{\mathrm{s} 1} ,則假定攝影機到受測者的距離 (CSD) 為 Delta D+D\Delta D+D ,而眼睛高度處的結構 S_(2)S_{2} (高度為 Deltah_(s1)+h_(s1)\Delta h_{s 1}+h_{s 1} )在視覺上會顯得與 S_(1)*S_(1)S_{1} \cdot S_{1} 相同大小,但由於放大失真,實際上會顯得比實際更大。請參閱圖 E.14。
Key 圖例
1 sensor plane 1 感測器平面
2 sensor 2 感測器
3 camera 3 攝影機
4 nose plane with S_(1)S_{1} object 4 個鼻平面與 S_(1)S_{1} 物件
5 eye plane with S_(2)S_{2} object 5 個眼平面與 S_(2)S_{2} 物件
6 optical axis 6 光軸
Figure E. 14 - Illustration of the magnification distortion effect 圖 E. 14 - 放大失真效應的圖示
For Figure E.14: (Deltah_(s1)+h_(s1))//h_(s1)=(Delta D+D)//D\left(\Delta h_{s 1}+h_{s 1}\right) / h_{s 1}=(\Delta D+D) / D. This leads to the definition of the magnification distortion factor: 對於圖 E.14: (Deltah_(s1)+h_(s1))//h_(s1)=(Delta D+D)//D\left(\Delta h_{s 1}+h_{s 1}\right) / h_{s 1}=(\Delta D+D) / D 。這導致了放大失真因子的定義:
the height of a structure S_(1)S_{1} in front of the face. 結構物在表面前方的高度 S_(1)S_{1} 。
Delta D is the depth of the measured object,
D+Delta D the camera-subject distance,
h_(s1) the height of a structure S_(1) in front of the face.| $\Delta D$ | is the depth of the measured object, |
| :--- | :--- |
| $D+\Delta D$ | the camera-subject distance, |
| $h_{s 1}$ | the height of a structure $S_{1}$ in front of the face. |
The magnification distortion factor K_("magnification ")K_{\text {magnification }} is the relative enlargement of an object at nose level compared to an object at eye level. The value of Delta D\Delta D is assumed to be 50 mm as the typical distance between the nose tip and the eye level of an adult Caucasian. Table E. 1 shows the (computed) absolute enlargement of an object with the size Delta D=50mm\Delta D=50 \mathrm{~mm} (like a nose) seen from several distances and the corresponding relative enlargements K_("magnification ")*K_{\text {magnification }} \cdot 放大失真係數 K_("magnification ")K_{\text {magnification }} 是鼻尖水平處的物件相對於眼睛水平處的物件的相對放大率。假設 Delta D\Delta D 的值為 50 毫米,這是成年白種人鼻尖與眼睛水平之間的典型距離。表 E.1 顯示了從不同距離觀察大小為 Delta D=50mm\Delta D=50 \mathrm{~mm} 的物件(例如鼻子)時的(計算出的)絕對放大率以及相對放大率 K_("magnification ")*K_{\text {magnification }} \cdot 。
Table E. 1 - Illustration of the magnification distortion effect 表 E.1 – 放大失真效應圖示
CSD in mm CSD (毫米)
Deltah_(s1)\Delta h_{s 1} in mm at h=50mm\boldsymbol{h}=\mathbf{5 0 ~ m m} Deltah_(s1)\Delta h_{s 1} 公釐,位於 h=50mm\boldsymbol{h}=\mathbf{5 0 ~ m m}
Figure E. 15 shows details of the photograph of the magnification distortion target taken in the intended photographic setup, from the chosen distance, with the intended focal length, and with the intended aperture. In order to determine the magnification distortion, measure a size h_(s1)+Deltah_(s1)h_{s 1}+\Delta h_{s 1} at nose-tip level and the difference to the corresponding size at eye level Deltah_(s1)\Delta h_{s 1}, and calculate 圖 E. 15 顯示了在預期攝影設定中,從選定距離、預期焦距和預期光圈拍攝的放大失真目標照片的詳細資訊。為了確定放大失真,請測量鼻尖高度的尺寸 h_(s1)+Deltah_(s1)h_{s 1}+\Delta h_{s 1} 以及與眼睛高度對應尺寸 Deltah_(s1)\Delta h_{s 1} 的差異,然後計算
Using the magnification distortion target from Figure E.10, the following experimental data have been collected. Figure E. 15 shows details from the captured image, Table E. 2 shows the measured and computed results. Note that the observations almost exactly match the computations from Table E.1. 使用圖 E.10 中的放大失真目標,已收集到以下實驗數據。圖 E.15 顯示了擷取影像的細節,表 E.2 顯示了測量和計算結果。請注意,觀測結果與表 E.1 中的計算結果幾乎完全吻合。
Figure E. 15 - Example measurement of the magnification distortion effect 圖 E.15 - 放大失真效應的測量範例
Table E. 2 - Experimental measures 表 E.2 – 實驗測量
CSD in mm CSD (毫米)
Target size h_(s1)+Deltah_(s1i)h_{s 1}+\Delta h_{s 1 i} in mm (and pixels) 目標尺寸 h_(s1)+Deltah_(s1i)h_{s 1}+\Delta h_{s 1 i} (單位:公釐,以及像素)
h_(s1)h_{s 1} in mm (and pixels) h_(s1)h_{s 1} (單位:公釐,以及像素)
Deltah_(s1)\Delta h_{s 1} in mm Deltah_(s1)\Delta h_{s 1} 公釐
Radial distortion of the camera lens causes information about the object to be misplaced but not lost. Camera lens measurement related radial distortion measurement is important as some of the ABC and photo kiosk cameras are not using highest quality lenses. For focal lengths shorter than about 30 mm , the distortion seems dominantly barrel, fish eye lens being the ultimate example. Cell phone cameras and fixed focus lenses have short focal lengths. 相機鏡頭的徑向畸變會導致物件資訊錯位,但不會遺失。相機鏡頭測量相關的徑向畸變測量很重要,因為有些 ABC 和相片快照亭的相機並未使用最高品質的鏡頭。對於焦距短於約 30 毫米的鏡頭,畸變似乎以桶狀畸變為主,魚眼鏡頭是最終極的例子。手機相機和定焦鏡頭的焦距都很短。
In practice measurements of the radial distortions are done either manually or using measurement software. If doing the measurements manually then it is important to zoom in the measured area and to use an image processing software ruler tool for measurements. In Figure E.16, the printed grid is marked above in light blue to show the recommended alignment and size when compared to a face portrait. Measurement vector starts from the grid corner point pixel location and ends at the respective corner point of the grid. The green circles are for reference purposes only to show how the recognition program may set certain landmarks on the actual face portrait. Guidance for the measurements is shown in Figure E.17. 在實務上,徑向失真的測量可透過手動或使用測量軟體來完成。如果手動測量,則務必放大測量區域,並使用影像處理軟體的尺規工具進行測量。在圖 E.16 中,上方以淺藍色標示的列印網格顯示了與臉部肖像相比時,建議的對齊方式和尺寸。測量向量從網格角點像素位置開始,並在網格的相應角點結束。綠色圓圈僅供參考,以顯示辨識程式如何在實際臉部肖像上設定某些地標。測量指南如圖 E.17 所示。
Figure E. 16 - Measurement grid example 圖 E.16 - 量測網格範例
Radial distortion is a type of geometrical aberration that causes a difference in magnification of the object at different points in the image. Various points are misplaced relative to the central point of the image. The barrel or pincushion distortion K_("radial ")K_{\text {radial }} is calculated by: 徑向畸變是一種幾何像差,會導致影像中不同點的物體放大率不同。各點相對於影像中心點的位置會錯位。桶形或枕形畸變 K_("radial ")K_{\text {radial }} 的計算方式為:
K_("radial ")=(PD-AD)//PD xx100K_{\text {radial }}=(P D-A D) / P D \times 100
where 其中 ADA D is the actual distance and ADA D 為實際距離,且 PDP D is the photographic distance from the centre point of the image. PDP D 為影像中心點的攝影距離。
Figure E. 17 gives an example of barrel distortion non-distorted and distorted frame intersection points shown with dots and arrows. Use guidelines to find the corner points of the undistorted image shown here as a grid drawn through the marked points with black lines. 圖 E.17 顯示桶狀畸變的範例,其中未畸變和畸變的畫面交點以點和箭頭表示。使用參考線找出未畸變影像的角點,此處以穿過標記點的黑色線條網格表示。
Figure E. 17 - Barrel distortion explained by a non-distorted and a distorted frame 圖 E.17 - 桶狀畸變的未畸變和畸變畫面解釋
Distortion, represented by a percentage, may be either positive or negative. A positive percentage represents pincushion distortion, whereas a negative percentage represents barrel distortion. 失真度以百分比表示,可能是正值或負值。正百分比表示枕狀失真,而負百分比表示桶狀失真。
Figure E. 18 illustrates barrel and pincushion distortion compared to an ideal grid picture in a perfectly square non-distorted image. 圖 E.18 說明了桶形和枕形失真與完美方形無失真影像中理想網格圖片的比較。
Figure E. 18 - Barrel and pincushion distortion 圖 E. 18 - 桶狀與枕狀畸變
In practice it is difficult to define the exact location of the non-distorted target location on the distorted photographic image. By drawing a cross through the middle point of the target image it is possible to find four intersecting points, which in turn may be used to locate the non-distorted target frame for error calculation purposes. This makes the measurements of different systems compatible, but the values are not absolute as the real distortion zooming error behaviour of different lenses are not the same. 實際上,很難在失真攝影影像上定義未失真目標的確切位置。透過在目標影像的中點畫一個十字,可以找到四個交點,這些交點又可用於定位未失真目標框,以進行誤差計算。這使得不同系統的測量結果具有相容性,但數值並非絕對,因為不同鏡頭的實際失真變焦誤差行為並不相同。
E. 3 Colour test E.3 色彩測試
E.3.1 Colour tests according to ISO/CIE 11664-4 E.3.1 依據 ISO/CIE 11664-4 的色彩測試
The IEC 61966-8 ^([11]){ }^{[11]}, ANSI IT8.7/2 ^([12]){ }^{[12]} or similar ISO 12641-1 ^([13]){ }^{[13]} test chart or other compatible and well documented colour test chart (e.g., IEC 61966-8 or ANSI IT8.7/2, see Figures E. 19 and E.20) containing skin colour patches should be used to measure colour accuracy, and dynamic range (indirect method). Colour quality is determined by converting the image to CIE L^(**)a^(**)b^(**[15])L^{*} a^{*} b^{*[15]} space and measuring the distance between the colours captured and known colour values of the chart patches. 應使用 IEC 61966-8 ^([11]){ }^{[11]} 、ANSI IT8.7/2 ^([12]){ }^{[12]} 或類似的 ISO 12641-1 ^([13]){ }^{[13]} 測試圖表或其他相容且文件齊全的色彩測試圖表(例如 IEC 61966-8 或 ANSI IT8.7/2,參見圖 E.19 和 E.20),其中包含膚色色塊,以測量色彩準確度和動態範圍(間接方法)。色彩品質是透過將影像轉換為 CIE L^(**)a^(**)b^(**[15])L^{*} a^{*} b^{*[15]} 空間,並測量所擷取色彩與圖表色塊已知色彩值之間的距離來確定的。
Figure E. 19 - IEC 61966-8 colour test chart 圖 E.19 - IEC 61966-8 色彩測試圖表
Patterns with sufficiently large patches are needed to measure noise (e. g., ISO-15739, ISO-14524). See Figure E.21. 需要具有足夠大色塊的圖案來測量雜訊(例如 ISO-15739、ISO-14524)。參見圖 E.21。
Figure E. 20 - ANSI IT8.7/2 test chart 圖 E.20 - ANSI IT8.7/2 測試圖
E.3.2 Image colour quality E.3.2 影像色彩品質
Human examiners and face recognition systems rely on high-quality skin tone presentation. Colour is a subjective psychological phenomenon, and human perception of colour depends on the context in which a perceived object is presented (i.e., chromatic adaptation). Therefore, the colour test should measure the entire gamut, as an examiner needs the surrounding colours to perceive colours in portions of the face (e.g., lips, hair, eyes, makeup) correctly. 人類檢查員和人臉辨識系統都依賴高品質的膚色呈現。色彩是一種主觀的心理現象,人類對色彩的感知取決於所感知物件呈現的環境(即色度適應)。因此,色彩測試應測量整個色域,因為檢查員需要周圍的色彩才能正確感知臉部各部分(例如嘴唇、頭髮、眼睛、妝容)的色彩。
Cameras should be white balanced and scanners colour managed to ensure high fidelity colour reproduction across the entire gamut. This is necessary as the digital camera software is preprocessing the internal raw image format and may distort the image colour when JPEG is used as source format for image analysis. 相機應進行白平衡,掃描器應進行色彩管理,以確保在整個色域中實現高保真色彩重現。這是必要的,因為數位相機軟體會預處理內部原始影像格式,並且當 JPEG 用作影像分析的來源格式時,可能會扭曲影像色彩。
In order to achieve good sample fidelity, there shall be no saturation (e.g., over- or under-exposure) on the measurement target. All RGB channels of the image should have at least 7 bits of intensity variation (i.e., span a range of at least 128 unique values) in the test target patch region of the image. This is required to get as near as possible to a L^(**)L^{*} level of 50 , which in turn ensures a wide sRGB ^([11]){ }^{[11]} gamut will be available for the analysis. 為了達到良好的樣本傳真度,測量目標上不應出現飽和(例如,過度曝光或曝光不足)。影像的所有 RGB 通道在影像的測試目標色塊區域中,應至少有 7 位元的強度變化(即,涵蓋至少 128 個獨特值的範圍)。這是為了盡可能接近 50 的 L^(**)L^{*} 水準,這反過來確保分析時能有寬廣的 sRGB ^([11]){ }^{[11]} 色域。
Figure E. 21 - Standard test patterns with sufficiently large patches to measure signal to noise 圖 E. 21 - 具有足夠大色塊以測量訊號雜訊比的標準測試圖案
E.3.3 Measurements and analysis E.3.3 測量與分析
In order to assure the required image quality system, installers shall make quality assurance measurements of light conditions and camera system response when a recommended CIE Standard Illuminant D65 high quality illuminant or similar continuous spectrum daylight illuminant and camera including camera control software are used to take pictures. In practice, it is also required to reduce the ambient light pollution emanating from uncontrolled daylight sources, fluorescent or similar light sources and reflections from surfaces. 為了確保所需的影像品質系統,安裝人員在使用推薦的 CIE 標準光源 D65 高品質光源或類似的連續光譜日光光源,以及包含相機控制軟體的相機進行拍攝時,應對光線條件和相機系統回應進行品質保證測量。實際上,也需要減少來自不受控制的日光光源、螢光燈或類似光源以及表面反射所產生的環境光污染。
Face portrait images are typically stored in sRGB IEC 61966-2-1[16], a device-independent colour space designed for consistent display across a wide range of commercial devices. However equal distances in spaces like sRGB do not represent equally perceptible differences between colour stimuli. To address this, in 1976 CIE created the LAB colour space whose coordinate system is based on nonlinear transformations which attempt to capture perceptual distance. When computing colour error, images are typically converted from the sRGB colour space to L^(**)a^(**)b^(**[15])\mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{*[15]}, a colour space engineered to approximate the way the human visual system perceives colour. Algorithmically, the transformation from sRGB to L^(**)a^(**)b^(**)\mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{*} values is accomplished by taking advantage of colour-matching functions developed for the CIE 1931 standard colourimetric system. 臉部肖像影像通常儲存在 sRGB IEC 61966-2-1[16] 中,這是一種與裝置無關的色彩空間,旨在各種商用裝置上實現一致的顯示。然而,sRGB 等空間中的等距離並不能代表色彩刺激之間同等可感知的差異。為了解決這個問題,CIE 於 1976 年創建了 LAB 色彩空間,其座標系統基於非線性轉換,試圖捕捉感知距離。在計算色彩誤差時,影像通常從 sRGB 色彩空間轉換為 L^(**)a^(**)b^(**[15])\mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{*[15]} ,這是一種旨在近似人類視覺系統感知色彩方式的色彩空間。在演算法上,從 sRGB 到 L^(**)a^(**)b^(**)\mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{*} 值的轉換是透過利用為 CIE 1931 標準色度系統開發的色彩匹配函數來實現的。
Fixed registration office imaging systems should be calibrated using manual or automated methods described in this document. For photo booth and mobile registration office imaging automatic white balance setting procedures and automatic quality analysis should be used. In a mobile environment, the use of advanced manual measurements may fail due to time and user training constraints. However, face portraits shall not be captured without adequate colour balance. 固定式註冊辦公室影像系統應使用本文件中描述的手動或自動方法進行校準。對於照相亭和行動註冊辦公室影像,應使用自動白平衡設定程序和自動品質分析。在行動環境中,由於時間和使用者訓練的限制,使用進階手動測量可能會失敗。然而,臉部肖像不應在沒有足夠色彩平衡的情況下拍攝。
Variations in human skin colour have been measured using visible reflectance spectroscopy and the device-independent colour space (CIELAB) ^([25]){ }^{[25]} [26]. Skin colour values can be expressed along the three dimensions of the CIELAB colour space: lightness scaled from 0 (black) to 100 (white) along L^(**)\mathrm{L}^{*}, and the opponent colour axes a^(**)\mathrm{a}^{*} and b^(**)\mathrm{b}^{*} representing from red through green from positive to negative values along a^(**)\mathrm{a}^{*} and similarly yellow through blue along b^(**)\mathrm{b}^{*}. 人類膚色的變化已使用可見光反射光譜和與裝置無關的色彩空間 (CIELAB) ^([25]){ }^{[25]} [26] 進行測量。膚色值可以沿著 CIELAB 色彩空間的三個維度表示:亮度沿著 L^(**)\mathrm{L}^{*} 從 0(黑色)到 100(白色)縮放,以及對立色彩軸 a^(**)\mathrm{a}^{*} 和 b^(**)\mathrm{b}^{*} ,分別表示沿著 a^(**)\mathrm{a}^{*} 從正值到負值的紅色到綠色,以及沿著 b^(**)\mathrm{b}^{*} 類似的黃色到藍色。
NOTE 1 When the face image is expressed in sRGB colour space, then the gamut shrinks and moves upwards (in a^(**)b^(**)\mathrm{a}^{*} \mathrm{~b}^{*} positive direction), and therefore higher L^(**)\mathrm{L}^{*} values may produce higher a^(**)\mathrm{a}^{*} and b^(**)\mathrm{b}^{*} values than shown in research papers. 註 1 當臉部影像以 sRGB 色彩空間表示時,色域會縮小並向上移動(在 a^(**)b^(**)\mathrm{a}^{*} \mathrm{~b}^{*} 正方向),因此較高的 L^(**)\mathrm{L}^{*} 值可能會產生比研究論文中顯示的更高的 a^(**)\mathrm{a}^{*} 和 b^(**)\mathrm{b}^{*} 值。
In a study[25] that used the above technique to measure the variation in the skin colour on the cheeks and foreheads of 960 Caucasian, Chinese, Kurdish, and Thai individuals, the expected mean and standard deviations of the variations found was as follows: L^(**)L^{*} mean 58,21 sigma 4,23 ; a* mean 11,45 sigma; 2,38 and b^(**)b^{*} mean 15,91 and sigma 2 . However, as the study population was not representative of the variation in the skin across all ethnicities, wider variations in skin colour should be expected. As 在一項研究[25]中,使用上述技術測量了 960 名白種人、中國人、庫德族和泰國人臉頰和前額的膚色變化,發現的預期平均值和標準差如下: L^(**)L^{*} 平均值 58.21,標準差 4.23;a*平均值 11.45,標準差 2.38;以及 b^(**)b^{*} 平均值 15.91,標準差 2。然而,由於研究人群不代表所有種族的皮膚變化,因此預計膚色會有更大的變化。
ISO/IEC 39794-5:2019(E)
studies[25][26] have shown that values lower than 5 for a^(**)\mathrm{a}^{*} and 10 for b^(**)\mathrm{b}^{*} have not been measured for face skin colours, these limits can be used to provide a warning of possible problems in captured skin colour. 研究[25][26]顯示,臉部膚色尚未測得低於 a^(**)\mathrm{a}^{*} 的 5 和低於 b^(**)\mathrm{b}^{*} 的 10 之數值,這些限制可用於對所擷取膚色中可能存在的問題發出警告。
NOTE 2 These studies did not take into account the possible impact on skin colour due to dermatological conditions. 附註 2 這些研究並未考量皮膚病學狀況對膚色可能造成的影響。
E.3.4 Coordinate calculation E.3.4 座標計算
Figure E. 22 - CIE chroma (ab) luminance (L*) levels 50 and 75 show the sRGB ^([11]){ }^{[11]} gamut compared to the entire a^(**)b^(**)\mathbf{a}^{*} \mathbf{b}^{*} area 圖 E.22 - CIE 彩度(ab)亮度(L*)等級 50 和 75 顯示了 sRGB ^([11]){ }^{[11]} 色域與整個 a^(**)b^(**)\mathbf{a}^{*} \mathbf{b}^{*} 區域的比較
Human skin tones should be located in the upper right hand sector of Figure E.22. 人類膚色應位於圖 E.22 的右上角區域。
In CIE Lab* the non-linear relations for L*, a* and b* are intended to mimic the logarithmic response of the human eye ^([14]){ }^{[14]}. Colour quality is determined by converting the image to CIEL^(**)a^(**)b^([15])\operatorname{CIE} \mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{[15]} space and measuring the distance between the colours captured and known colour values of the chart patches. CIE Delta E 2000[23] is a standard method for measuring this distance. A capture system’s performance can be improved by minimizing the system’s average (i.e., measured across all chart patches) and maximum (i.e., measured for any given chart patch) Delta E 2000. 在 CIE Lab*中,L*、a*和 b*的非線性關係旨在模仿人眼的對數回應 ^([14]){ }^{[14]} 。色彩品質是透過將影像轉換為 CIEL^(**)a^(**)b^([15])\operatorname{CIE} \mathrm{L}^{*} \mathrm{a}^{*} \mathrm{~b}^{[15]} 空間並測量所捕捉到的顏色與圖表色塊已知顏色值之間的距離來確定的。CIE Delta E 2000[23]是測量此距離的標準方法。捕捉系統的效能可以透過最小化系統的平均(即,測量所有圖表色塊)和最大(即,測量任何給定圖表色塊)Delta E 2000 來提高。
An ideal system would have an average deltaE2000 of 1 and a maximum deltaE2000 of 5 . 一個理想的系統,其平均 deltaE2000 為 1,最大 deltaE2000 為 5。
E. 4 MTF test method according to ISO 12233:2014 E. 4 依據 ISO 12233:2014 的 MTF 測試方法
Spatial resolution is a measure of the smallest discernible detail in an image. The image resolution measurements described herein are designed for the calibration of photo studios and office imaging systems. The methods generally involve photographing or scanning a standard target and analysing the resulting images on a computer using standardized algorithms to compute a value. 空間解析度是衡量影像中可辨識最小細節的指標。本文所述的影像解析度測量旨在校準攝影工作室和辦公室影像系統。這些方法通常涉及拍攝或掃描標準目標,並使用標準化演算法在電腦上分析所得影像以計算數值。
Image fidelity factors are affected by the imaging sensor and lens. Resolution is a single frequency parameter that indicates whether the output signal contains a minimum threshold of detail information for visual detection (i.e., the highest spatial frequency that a camera or other similar imaging device can usefully capture). 影像保真度因素受成像感測器和鏡頭的影響。解析度是一個單頻參數,表示輸出訊號是否包含用於視覺偵測的最小細節資訊閾值(即相機或其他類似成像裝置可有效捕捉的最高空間頻率)。
Figure E. 23 - Relation between the physical size of an object at the face and its counterpart at the camera sensor 圖 E. 23 - 物件在臉部實體尺寸與其在相機感測器上對應部分的關係
Spatial frequency response is a multi-valued metric that measures contrast loss as a function of spatial frequency. Generally, contrast decreases as a function of spatial frequency to a level where detail can no longer be visually resolved. See Figure E.23. This limiting frequency value is the resolution of the camera, which is determined by the performance of the camera lens, the number of addressable photo elements in the optical imaging device, and the electrical circuits in the camera, which optionally perform image compression and gamma correction. 空間頻率響應是一種多值度量,用於測量對比度損失與空間頻率的函數關係。通常,對比度會隨著空間頻率的增加而降低,直到細節無法再透過視覺辨識。請參閱圖 E.23。此限制頻率值即為相機的解析度,其由相機鏡頭的性能、光學成像裝置中可定址光電元件的數量,以及相機中可選擇性執行影像壓縮和伽瑪校正的電路所決定。
Figure E. 24 - Sine wave test chart overlaid with a white contrast mask to show decreasing contrast effects from 100%100 \% value at the top to a 0%0 \% value at the bottom 圖 E.24 - 弦波測試圖疊加白色對比遮罩,顯示對比度從頂部的 100%100 \% 值到底部的 0%0 \% 值逐漸降低的效果
Modulation transform functions (MTFs) are normally measured by optics experts using purposeprinted sine wave test charts, such as in Figure E.24, and standardized procedures. Alternatively, MTF can be determined from the magnitude of the Fourier transform of a system’s point or line spread function. The Fourier transform decomposes the spread function into the frequencies that comprise it. By looking at the amplitudes of the frequencies, it is possible to define the resolution characteristics of the measured imaging system. 調變轉換函數 (MTF) 通常由光學專家使用專門印製的弦波測試圖(如圖 E.24 所示)和標準化程序進行測量。或者,MTF 也可以透過系統點或線擴散函數的傅立葉轉換幅度來確定。傅立葉轉換將擴散函數分解為構成它的頻率。透過觀察頻率的振幅,可以定義所測量成像系統的解析度特性。
Figure E. 25 - Test pattern: ISO 12233:2014 圖 E. 25 - 測試圖案:ISO 12233:2014
ISO 12233:2014 specifies methods for measuring the spatial frequency response (SFR) of electronic still picture cameras and similar imaging devices such as video cameras and flatbed scanners. The SFR measurement closely approximates the mathematically-defined system MTF of the camera. The MTF of the camera can only be approximated through the SFR, because most electronic still-picture cameras provide spatial colour sampling and nonlinear processing. ISO 12233:2014 規定了測量電子靜態影像相機和類似成像裝置(如攝影機和平台式掃描器)空間頻率響應 (SFR) 的方法。SFR 測量與相機的數學定義系統 MTF 非常接近。相機的 MTF 只能透過 SFR 近似,因為大多數電子靜態影像相機提供空間色彩取樣和非線性處理。
SFR is measured by capturing an image of a bi-tonal, rotated square test pattern, such as those shown in Figure E. 25 or E.26, which can subsequently be analysed using readily-available edge analysis image processing software. It is important that the test pattern is captured in the same photographic environment that is prescribed for the face by this document, including camera-to-subject distance, image dimensions, etc. SFR 的測量方式是拍攝雙色、旋轉方形測試圖案的影像,例如圖 E.25 或 E.26 所示的圖案,然後可以使用現成的邊緣分析影像處理軟體進行分析。重要的是,測試圖案必須在與本文件規定的人臉拍攝環境相同的攝影環境中拍攝,包括相機到主體的距離、影像尺寸等。
To perform slanted edge measurements using a software program, the user selects a region containing the edge to be measured. The software processes digital image values near slanted vertical and horizontal black to white edges to derive super sampled edge spread data, which is then filtered and converted to the frequency domain to get the SFR values. Horizontal edge is used for vertical SFR measurement. 若要使用軟體程式執行傾斜邊緣測量,使用者會選取包含要測量邊緣的區域。軟體會處理傾斜垂直和水平黑白邊緣附近的數位影像值,以導出超取樣邊緣擴散資料,然後將其濾波並轉換為頻域以取得 SFR 值。水平邊緣用於垂直 SFR 測量。
Slanted edge measurements are less sensitive to noise than sine patterns. Gamma influences the MTF measurement accuracy, and for this reason, gamma value should be measured using a grey scale chart. Gamma correction, or often simply gamma, is the name of a nonlinear operation used to encode and decode luminance or colour related tristimulus values in imaging systems. An incorrect gamma setting for the MTF calculation causes an error situation where the MTF at 2 * Nyquist is not equal to 0 , as it should be. In practice, a measurement of exactly 0 is not required to achieve an acceptable measurement. 傾斜邊緣測量對雜訊的敏感度低於正弦圖案。伽瑪會影響 MTF 測量準確度,因此應使用灰階圖表測量伽瑪值。伽瑪校正,或通常簡稱為伽瑪,是一種非線性操作的名稱,用於在成像系統中編碼和解碼亮度或顏色相關的三刺激值。MTF 計算的伽瑪設定不正確會導致錯誤情況,即 2 * 奈奎斯特頻率下的 MTF 不等於 0,而它應該等於 0。實際上,不需要精確測量到 0 即可實現可接受的測量。
Figure E. 26 - Test pattern: ISO 16067-1 圖 E. 26 - 測試圖案:ISO 16067-1
Note that the physical distance between the fiducials on the standardized test pattern is 66,8mm66,8 \mathrm{~mm}. 請注意,標準化測試圖案上參考點之間的實際距離為 66,8mm66,8 \mathrm{~mm} 。
An example SFR measured for an imaging system is shown in Figure E.27. The contrast value is depicted either in percent or on a scale from 0 to 1 , where 1 corresponds to 100%100 \% on the vertical axis, and frequency values start from 0 and increase to the right on the horizontal axis. The frequency unit of an SFR measurement can be represented in cy//px\mathrm{cy} / \mathrm{px} or cy//mm\mathrm{cy} / \mathrm{mm}. For the purposes of this document, where the system specification references the object being photographed (i. e., the size of features on a face), the frequency unit should be in cycles //mm/ \mathrm{mm} in the object plane. The size of a freckle/mole that should be detectable in face photos is 2 mm to 3 mm . The unit cy/ mm is preferred over cy/px as sensor dimensions, camera distances, and head sizes vary. The physical distance of the object is the universal unit for all issuers. This measurement methodology requires that the user compute the sampling rate in px//mm\mathrm{px} / \mathrm{mm}, which is determined by measuring the number of pixels across the known physical dimensions of the test pattern. 影像系統的 SFR 測量範例顯示於圖 E.27。對比值以百分比或 0 到 1 的比例表示,其中 1 對應於垂直軸上的 100%100 \% ,頻率值從 0 開始並向右增加於水平軸上。SFR 測量的頻率單位可以表示為 cy//px\mathrm{cy} / \mathrm{px} 或 cy//mm\mathrm{cy} / \mathrm{mm} 。就本文件而言,當系統規格參考被拍攝的物件(即臉部特徵的大小)時,頻率單位應為物件平面中的每 //mm/ \mathrm{mm} 循環。臉部照片中應可偵測到的雀斑/痣大小為 2 毫米至 3 毫米。相較於 cy/px,cy/mm 單位更受青睞,因為感測器尺寸、相機距離和頭部大小各不相同。物件的物理距離是所有發行者的通用單位。此測量方法要求使用者計算每 px//mm\mathrm{px} / \mathrm{mm} 的取樣率,該取樣率是透過測量測試圖案已知物理尺寸上的像素數量來確定的。
Key
1 Nyquist 鍵 1 奈奎斯特
Figure E. 27 - Example SFR graph 圖 E. 27 - SFR 圖範例
The Nyquist frequency is half of the sampling rate of a discrete (opposite of continuous) signal processing system. The Nyquist frequency is also called the Nyquist limit. It is the highest frequency that can be encoded at a given sampling rate while still fully reconstructing the image. The Nyquist frequency for an imaging system is 0,5cy//px0,5 \mathrm{cy} / \mathrm{px} as two samples (i.e., pixels) is the minimum required to represent a complete cycle. However, cy/px does not clarify the size of the object being resolved and must therefore be converted to cy/mm. 奈奎斯特頻率是離散(與連續相對)訊號處理系統取樣率的一半。奈奎斯特頻率也稱為奈奎斯特極限。它是給定取樣率下可以編碼的最高頻率,同時仍能完全重建影像。成像系統的奈奎斯特頻率為 0,5cy//px0,5 \mathrm{cy} / \mathrm{px} ,因為兩個取樣(即像素)是表示完整週期所需的最小值。然而,cy/px 並未闡明所解析物件的大小,因此必須轉換為 cy/mm。
The SFR is deemed acceptable if the MTF remains sufficiently high up to a specified frequency. The Nyquist limit for the specification is determined by the minimum sampling rate specified in this document multiplied by the Nyquist limit in cy/px. A different specification is needed for scanned photographs versus live capture, as the object plane in each case is different. The minimum sampling rates in this document are: 90 pixels per inter-eye distance (approximately 60 mm ) for cameras, and 300 PPI for scanners. The Nyquist limits are therefore: 0,75cy//mm0,75 \mathrm{cy} / \mathrm{mm} for cameras and 5,9cy//mm5,9 \mathrm{cy} / \mathrm{mm} for scanners. 如果 MTF 在達到指定頻率前仍保持足夠高,則 SFR 被視為可接受。本文件中規定的最小取樣率乘以 cy/px 中的奈奎斯特極限,即可確定該規範的奈奎斯特極限。掃描照片與即時擷取需要不同的規範,因為兩者的物體平面不同。本文件中的最小取樣率為:相機為每眼距 90 像素(約 60 毫米),掃描器為 300 PPI。因此,奈奎斯特極限為:相機為 0,75cy//mm0,75 \mathrm{cy} / \mathrm{mm} ,掃描器為 5,9cy//mm5,9 \mathrm{cy} / \mathrm{mm} 。
MTF20, which is an indicator of SFR at higher spatial frequencies, should occur at approximately 80%80 \% of the Nyquist frequency, or 0,6cy//mm0,6 \mathrm{cy} / \mathrm{mm} for cameras and 4,7cy//mm4,7 \mathrm{cy} / \mathrm{mm} for scanners. MTF20 是較高空間頻率下 SFR 的指標,應約為奈奎斯特頻率的 80%80 \% ,或相機為 0,6cy//mm0,6 \mathrm{cy} / \mathrm{mm} ,掃描器為 4,7cy//mm4,7 \mathrm{cy} / \mathrm{mm} 。
E. 5 Focus and depth-of-field considerations E. 5 對焦與景深考量
Proper focus and depth-of-field will be assured by pre-focusing the lens at the distance of the subject’s eyes and by selecting an appropriate aperture ( F -stop) to ensure a depth-of-field containing a subject’s nose and ears. The depth-of-field of a lens is dependent upon its focal length, its effective aperture, and the focus distance. Point sources which are closer or farther than the distance at which a lens is well focused will be blurred, with the extent of the blur described by a “circle of confusion”. If the maximum diameter of the circle of confusion is limited by, for example, the spacing between adjacent pixels in a CCD image sensor, the front and rear distances from the plane of optimum focus that produce acceptably focused images can be determined. The sum of these front and rear distances is the depth-of-field. 透過將鏡頭預先對焦在拍攝對象眼睛的距離,並選擇適當的光圈(F 值)以確保景深包含拍攝對象的鼻子和耳朵,即可確保適當的對焦和景深。鏡頭的景深取決於其焦距、有效光圈和對焦距離。比鏡頭良好對焦距離更近或更遠的點光源將會模糊,模糊的程度由「模糊圈」描述。如果模糊圈的最大直徑受到限制,例如,CCD 影像感測器中相鄰像素之間的間距,則可以確定產生可接受對焦影像的最佳對焦平面前後的距離。這些前後距離的總和即為景深。
D_("rear ")=(cFs(s-f))/(f^(2)-cF(s-f))D_{\text {rear }}=\frac{c F s(s-f)}{f^{2}-c F(s-f)}
where 其中
D_("DoF ")D_{\text {DoF }}
is the depth of field, 是景深,
D_("front ")D_{\text {front }}
is the front focal distance, the distance from the plane of focus to the plane closest to the lens that is still in acceptable focus, 是前焦距,即從對焦平面到離鏡頭最近且仍在可接受對焦範圍內的平面的距離,
D_("rear ")D_{\text {rear }}
is the rear focal distance, the distance from the plane of focus to the plane farthest from the lens that is still in acceptable focus, 是後焦距,即從對焦平面到離鏡頭最遠且仍在可接受對焦範圍內的平面的距離,
cc
is the diameter of the circle of confusion, 是模糊圈的直徑,
ss
is the distance from the lens to the object, and 是鏡頭到物件的距離,以及
F=f//aF=f / a
is the F-stop, the lens focal length ff divided by the effective lens aperture aa. 是光圈值,即鏡頭焦距 ff 除以有效鏡頭光圈 aa 。
D_("DoF ") is the depth of field,
D_("front ") is the front focal distance, the distance from the plane of focus to the plane closest to the lens that is still in acceptable focus,
D_("rear ") is the rear focal distance, the distance from the plane of focus to the plane farthest from the lens that is still in acceptable focus,
c is the diameter of the circle of confusion,
s is the distance from the lens to the object, and
F=f//a is the F-stop, the lens focal length f divided by the effective lens aperture a.| $D_{\text {DoF }}$ | is the depth of field, |
| :--- | :--- |
| $D_{\text {front }}$ | is the front focal distance, the distance from the plane of focus to the plane closest to the lens that is still in acceptable focus, |
| $D_{\text {rear }}$ | is the rear focal distance, the distance from the plane of focus to the plane farthest from the lens that is still in acceptable focus, |
| $c$ | is the diameter of the circle of confusion, |
| $s$ | is the distance from the lens to the object, and |
| $F=f / a$ | is the F-stop, the lens focal length $f$ divided by the effective lens aperture $a$. |
Figure E. 28 illustrates these dimensions. 圖 E.28 說明了這些尺寸。
Figure E. 28 - Dimensions for depth-of-field calculations 圖 E.28 - 景深計算的尺寸
E. 6 Report about the study of the effect of the camera subject distance of reference face images on face verification performances E.6 關於參考臉部影像的相機主體距離對臉部驗證效能影響之研究報告
E.6.1 What is magnification distortion? E.6.1 什麼是放大失真?
Taking photographs from a short camera subject distance causes magnification distortion of face images. Figure E. 14 illustrates the magnification distortion. Let the distance between the tip of the nose of a capture subject and the lens of a camera be DD, the distance between the eyes plane of the capture subject and the lens of the camera, i. e. the camera-subject distance, be D+Delta DD+\Delta D, and the height of a structure at nose level be hh. Then, the structure at nose level appears to have the same size as a structure of the height h+Delta hh+\Delta h at eye level. The magnification distortion is defined as: 從短距離拍攝照片會導致臉部影像的放大失真。圖 E.14 說明了放大失真。設拍攝對象的鼻尖與相機鏡頭之間的距離為 DD ,拍攝對象的眼睛平面與相機鏡頭之間的距離(即相機與對象之間的距離)為 D+Delta DD+\Delta D ,以及鼻部高度的結構為 hh 。那麼,鼻部高度的結構看起來與眼睛高度為 h+Delta hh+\Delta h 的結構大小相同。放大失真定義為:
If the distance between the nose plane and the eyes plane Delta D\Delta D is 50 mm , the magnification distortions are as in Table E.3. 如果鼻平面與眼平面之間的距離為 50 mm,則放大失真如表 E.3 所示。
Table E. 3 - Magnification distortion as a function of camera-subject distance 表 E.3 – 放大失真與相機-主體距離的函數關係
E.6.2 Methodology from enrolment to score calculation E.6.2 從註冊到分數計算的方法論
A bench has been created capable of rapidly capturing pictures of the same subject under strictly controlled conditions. That bench (see Figure E.29) ensured that for all face images, the capture conditions were the same and in line with the provisions specified in this document, except for the camera subject distance, which is variable and ranges from 0,5m0,5 \mathrm{~m} to 3 m . The ten distances are 0,5m0,5 \mathrm{~m}, 0,6m,0,7m,0,8m,0,9m,1,0m,1,5m,2,0m,2,5m0,6 \mathrm{~m}, 0,7 \mathrm{~m}, 0,8 \mathrm{~m}, 0,9 \mathrm{~m}, 1,0 \mathrm{~m}, 1,5 \mathrm{~m}, 2,0 \mathrm{~m}, 2,5 \mathrm{~m} and 3,0m3,0 \mathrm{~m}. 已建立一個能夠在嚴格控制的條件下快速捕捉同一主體影像的測試台。該測試台(參見圖 E.29)確保所有臉部影像的捕捉條件均相同,並符合本文件所規定的條款,但相機與主體之間的距離除外,該距離是可變的,範圍從 0,5m0,5 \mathrm{~m} 到 3 公尺。這十個距離是 0,5m0,5 \mathrm{~m} 、 0,6m,0,7m,0,8m,0,9m,1,0m,1,5m,2,0m,2,5m0,6 \mathrm{~m}, 0,7 \mathrm{~m}, 0,8 \mathrm{~m}, 0,9 \mathrm{~m}, 1,0 \mathrm{~m}, 1,5 \mathrm{~m}, 2,0 \mathrm{~m}, 2,5 \mathrm{~m} 和 3,0m3,0 \mathrm{~m} 。
The bench contains a Canon EOS 6D digital camera: 測試台包含一台 Canon EOS 6D 數位相機:
The lens used is a Canon EF 50mmf//1,8STM50 \mathrm{~mm} \mathrm{f} / 1,8 \mathrm{STM} with radial distortion of less than 0,8%0,8 \%. The bench uses two front flashes PROFILITE 250 and one background flash PROFILITE 250 (each 250 W ). The two front flashes are set to 5,0 . The background flash is set to 1,0 . The ambient light of the flashes is set to 50%50 \%. 所使用的鏡頭為 Canon EF 50mmf//1,8STM50 \mathrm{~mm} \mathrm{f} / 1,8 \mathrm{STM} ,其徑向畸變小於 0,8%0,8 \% 。測試台使用兩盞前置閃光燈 PROFILITE 250 和一盞背景閃光燈 PROFILITE 250(每盞 250 瓦)。兩盞前置閃光燈設定為 5.0。背景閃光燈設定為 1.0。閃光燈的環境光設定為 50%50 \% 。
After seating a test subject in front of the camera of the bench, a capture session is started. The camera automatically captures 40 face images in two passes. In each pass, the camera moves to ten positions with different camera-subject distances from 0,5m0,5 \mathrm{~m} to 3 m . In the first pass, the camera moves away from the test subject. In the second pass, the camera moves back towards the test subject. At each stop, two images are taken in order to mitigate the risk of closed-eye effect due to flash. After capturing 40 images, the capture session is completed. 在測試對象坐在測試台的攝影機前之後,便會開始一個擷取工作階段。攝影機會自動分兩次擷取 40 張臉部影像。在每次擷取中,攝影機會移動到十個位置,這些位置的攝影機與受測者距離從 0 到 3 公尺不等。在第一次擷取中,攝影機會遠離受測者。在第二次擷取中,攝影機會移回受測者。在每個停靠點,都會拍攝兩張影像,以減輕因閃光燈造成的閉眼效應風險。擷取 40 張影像後,擷取工作階段便會完成。
Time between acquisitions is 12 s , the total duration of a capture session is 4 min . The precision of the movement is below 2 mm ( 0,08%0,08 \% of the full movement). Given the difficulty to locate the optical centre of the camera lens and given the morphological and behavioural differences of the test subjects, the actual camera-subject distances may be up to 30 mm smaller than the recorded camera subject distances. 每次擷取之間的時間為 12 秒,一個擷取工作階段的總持續時間為 4 分鐘。移動的精確度低於 2 毫米(佔總移動的 0)。考量到攝影機鏡頭光學中心的定位困難,以及受測者的形態和行為差異,實際的攝影機與受測者距離可能比記錄的攝影機與受測者距離小 30 毫米。
Figure E. 29 - Bench created to rapidly capture a number of pictures at different distances 圖 E. 29 – 為快速擷取不同距離的多張影像而設計的測試台
Using the bench, local databases of face images with different camera subject distances were collected at the premises of members of the study group from as many volunteer test subjects as possible. 利用此測試台,研究小組成員在各自的場所收集了盡可能多志願受試者在不同攝影機與受試者距離下的臉部影像本地資料庫。
Local databases were collected at the premises of (in temporal order): 本地資料庫的收集地點(依時間順序)為:
KIS SAS in France, 法國的 KIS SAS,
Oberthur Technologies in France, 法國的歐貝特科技(Oberthur Technologies),
Photo-Me International in the UK, 英國的快照我國際(Photo-Me International),
Gemalto in the Czech Republic, 捷克的金雅拓(Gemalto),
Fotofix Schnellphotoautomaten in Germany, and 以及德國的 Fotofix Schnellphotoautomaten,
Nippon Auto-Photo in Japan. 日本 Nippon Auto-Photo。
Each test subject participated in only one capture session. The captured face images were cropped and resized in conformity with this document (i.e. “ICAO cropped”). The format of the ICAO cropped images was the JPEG file interchange format. 每位受測者僅參與一次擷取工作階段。擷取的臉部影像依據本文件進行裁切和調整大小(即「ICAO 裁切」)。ICAO 裁切影像的格式為 JPEG 檔案交換格式。
The local databases were encrypted and then sent to the Biometrics Evaluation Laboratory at the Fraunhofer Institute for Computer Graphics Research IGD, in order to be processed using various state-of-the-art face recognition algorithms. 本地資料庫經過加密後,傳送至 Fraunhofer 電腦圖形研究中心 IGD 的生物辨識評估實驗室,以便使用各種最先進的臉部辨識演算法進行處理。
The local databases were merged into one consolidated database containing 20 ICAO cropped face images from 435 test subjects, i.e. in total 8700 images. The filenames of all face images were pseudonymised such that it was not apparent from the filenames which face images matched and from which camera-subject distance they were captured. The consolidated face image database was divided into a directory of reference face images and a directory of probe face images. For each test subject, the directory of reference images contained ten ICAO cropped face images from the first pass (one per camera-subject distance), and the directory of probe images contained ten ICAO cropped face images from the second pass (one per camera-subject distance). The consolidated face image database was sequestered for the official run of face comparisons. 本地資料庫合併為一個整合式資料庫,其中包含來自 435 名受測者的 20 張 ICAO 裁切臉部影像,總計 8700 張影像。所有臉部影像的檔案名稱均經過假名化處理,因此無法從檔案名稱判斷哪些臉部影像相符,以及它們是從哪個相機與受測者距離擷取而來。整合式臉部影像資料庫分為參考臉部影像目錄和探測臉部影像目錄。對於每位受測者,參考影像目錄包含第一次拍攝的十張 ICAO 裁切臉部影像(每個相機與受測者距離一張),探測影像目錄包含第二次拍攝的十張 ICAO 裁切臉部影像(每個相機與受測者距離一張)。整合式臉部影像資料庫已封存,以供官方進行臉部比對。
ISO/IEC 39794-5:2019(E)
Providers of commercial off-the-shelf, state-of-the-art algorithms for one-to-one face comparison were invited to participate in this study. The following algorithm providers (in alphabetic order) submitted face comparison software: 本研究邀請了提供現成、最先進一對一臉部比對演算法的廠商參與。以下演算法供應商(依字母順序排列)提交了臉部比對軟體:
Dermalog, Dermalog,
id3 Technologies, id3 Technologies,
Innovatrics,
NEC, and NEC,以及
OT-Morpho. OT-Morpho。
The executable face comparison software was submitted by the algorithm providers to Fraunhofer IGD to be executed there on the sequestered face image database. 可執行的人臉比對軟體由演算法供應商提交給 Fraunhofer IGD,以便在隔離的人臉影像資料庫上執行。
Each participating face comparison algorithm has been interfaced with two software interfaces. The participating algorithm providers were asked to build two executables in the form of Windows console applications: 每個參與的人臉比對演算法都已與兩個軟體介面連接。參與的演算法供應商被要求建置兩個可執行檔,形式為 Windows 控制台應用程式:
extract.exe, to extract comparable features from each of a list of face images, and extract.exe,用於從臉部影像清單中的每個影像中提取可比較的特徵,以及
compare.exe, to compare features and produce a comparison score for each pair of faces given in a comparison list. compare.exe,用於比較特徵並為比較清單中給定的每對臉部產生比較分數。
Each participating face comparison algorithm compared the features from each reference face image with the features from each probe face image. This means 4350 xx4350=189225004350 \times 4350=18922500 comparisons. For each comparison the file name of the reference image, the file name of the probe image, and the comparison score has been recorded in a CSV file. 每個參與的人臉比對演算法會比對來自每張參考人臉影像的特徵與來自每張探測人臉影像的特徵。這表示有 4350 xx4350=189225004350 \times 4350=18922500 次比對。對於每次比對,參考影像的檔案名稱、探測影像的檔案名稱以及比對分數都已記錄在 CSV 檔案中。
E.6.3 Data analysis E.6.3 資料分析
E.6.3.1 Methodology E.6.3.1 方法論
The research hypothesis is that: 研究假設為:
the camera-subject distance of a face image or 人臉影像的攝影機與受測者距離,或
different magnification distortions of a face image and of probe face images compared with that face image 與該臉部影像相比,臉部影像和探測臉部影像的不同放大失真
have an effect on the usefulness of that face image as a reference image. The usefulness of biometric sample for telling mated and non-mated samples apart is referred to as “utility”[27]. If for several of the participating state-of-the-art face comparison algorithms, the camera-subject distance of reference images has negligible effect on their utility, the research hypothesis would be refuted. 會影響該臉部影像作為參考影像的實用性。生物識別樣本在區分配對和非配對樣本方面的實用性被稱為「效用」[27]。如果對於幾個參與的尖端臉部比對演算法,參考影像的相機與受試者距離對其效用影響可忽略不計,則研究假設將被駁斥。
E.6.3.2 False non-match rate at fixed false match rate E.6.3.2 在固定錯誤比對率下的錯誤不比對率
For four out of five of the participating algorithms, the highest non-mated similarity score is lower than the lowest mated similarity score, i. e. the distribution of mated scores was clearly separated from that of non-mated scores, allowing perfect classification by setting the decision threshold between the two distributions of scores. No matter what is the allowed value for FMR > 0%0 \%, no false non-match error was observed in 43500 mated comparisons. Thus, according to the "Rule of 3 "[28], with 95%95 \% confidence FNMR <= 0,0069%\leq 0,0069 \% for most of the participating algorithms. 對於五分之四的參與演算法,最高非配對相似度分數低於最低配對相似度分數,亦即配對分數的分佈與非配對分數的分佈明顯分離,透過將決策閾值設定在兩個分數分佈之間,可以實現完美的分類。無論 FMR > 0%0 \% 的允許值為何,在 43500 次配對比較中,未觀察到任何錯誤的非配對錯誤。因此,根據「三法則」[28],對於大多數參與演算法,FNMR <= 0,0069%\leq 0,0069 \% 的置信度為 95%95 \% 。
E.6.3.3 Distance-related differences in score distributions E.6.3.3 分數分佈中與距離相關的差異
Distance-related differences in score distributions can be significant in real scenarios where mated similarity scores are lower because of other factors affecting recognition performance (such as aging, pose variation and illumination). 在實際情境中,由於其他影響辨識效能的因素(例如老化、姿勢變化和光線),配對相似度分數較低,因此分數分佈中與距離相關的差異可能很顯著。
A measure of how well the distributions of mated and non-mated comparison scores are separated is d^(')d^{\prime} (pronounced “d-prime”), defined as d^(')=(|mu_(m)-mu_(n)|)/(sqrt(sigma_(m)^(2)+sigma_(n)^(2)))d^{\prime}=\frac{\left|\mu_{m}-\mu_{n}\right|}{\sqrt{\sigma_{m}{ }^{2}+\sigma_{n}{ }^{2}}}, where 衡量配對與非配對比較分數分佈分離程度的指標是 d^(')d^{\prime} (發音為「d-prime」),定義為 d^(')=(|mu_(m)-mu_(n)|)/(sqrt(sigma_(m)^(2)+sigma_(n)^(2)))d^{\prime}=\frac{\left|\mu_{m}-\mu_{n}\right|}{\sqrt{\sigma_{m}{ }^{2}+\sigma_{n}{ }^{2}}} ,其中
mu_(m)\mu_{m} is the arithmetic mean of the mated comparison scores, mu_(m)\mu_{m} 是配對比較分數的算術平均值,
mu_(n)\mu_{n} is the arithmetic mean of the non-mated comparison scores, mu_(n)\mu_{n} 是非比對比較分數的算術平均值,
sigma_(m)\sigma_{m} is the standard deviation of the mated comparison scores, and sigma_(m)\sigma_{m} 是配對比較分數的標準差,且
sigma_(n)\sigma_{n} is the standard deviation of the non-mated comparison scores[29]. sigma_(n)\sigma_{n} 是非比對分數的標準差[29]。
Figure E. 30 shows the average d^(')d^{\prime} values over the three best participating commercial face comparison algorithms as a function of the camera-subject distance of the reference image and the camera-subject distance of the probe image. The individual values are represented as colours. The lowest value is mapped to dark blue and the highest value to dark red. 圖 E.30 顯示了三個最佳參與商業人臉比對演算法的平均 d^(')d^{\prime} 值,其為參考影像的相機與受試者距離以及探測影像的相機與受試者距離的函數。個別值以顏色表示。最低值對應深藍色,最高值對應深紅色。
Figure E. 30 - Average d^(')\boldsymbol{d}^{\prime} values over three commercial face comparison algorithms 圖 E. 30 - 三種商用人臉比對演算法的平均 d^(')\boldsymbol{d}^{\prime} 值
E.6.4 Theoretical predictions E.6.4 理論預測
If a reference face image is compared with NN face images, an algorithm returns NN scores d_(1)d_{1} to s_(N)s_{\mathrm{N}}. We sort these NN values in a decreasing order to get s_(1)^(')s_{1}^{\prime} to s_(N)^(')(s_(1)^('):}s_{N}^{\prime}\left(s_{1}^{\prime}\right. is the biggest score). Considering a d^(')d^{\prime} value over 12 means that the probability to have s_(1)^(')s_{1}^{\prime} not corresponds to the right individual is below 2,15*10^(-32)2,15 \cdot 10^{-32} (see the demonstration below). 如果將參考臉部影像與 NN 個臉部影像進行比較,演算法會傳回 NN 個分數 d_(1)d_{1} 到 s_(N)s_{\mathrm{N}} 。我們將這些 NN 值依遞減排序,以取得 s_(1)^(')s_{1}^{\prime} 到 s_(N)^(')(s_(1)^('):}s_{N}^{\prime}\left(s_{1}^{\prime}\right. (其中 s_(N)^(')(s_(1)^('):}s_{N}^{\prime}\left(s_{1}^{\prime}\right. 是最大分數)。考量到 d^(')d^{\prime} 值超過 12 意味著 s_(1)^(')s_{1}^{\prime} 不符合正確個體的機率低於 2,15*10^(-32)2,15 \cdot 10^{-32} (請參閱下面的示範)。
If N is the world population of 7*10^(9)7 \cdot 10^{9} people, we estimate the probability to have one individual wrongly classified due to magnification distortion extrapolates to 77*10^(9)*2,15*10^(-32)=1,51*10^(-22)77 \cdot 10^{9} \cdot 2,15 \cdot 10^{-32}=1,51 \cdot 10^{-22}. This estimate assumes the comparison scores are normally distributed, that the mated pairs are captured in 如果 N 是全球人口數 7*10^(9)7 \cdot 10^{9} ,我們估計因放大失真而錯誤分類的個體機率外推至 77*10^(9)*2,15*10^(-32)=1,51*10^(-22)77 \cdot 10^{9} \cdot 2,15 \cdot 10^{-32}=1,51 \cdot 10^{-22} 。此估計假設比對分數呈常態分佈,且配對樣本是在擷取時
ISO/IEC 39794-5:2019(E)
a single sitting, and that the comparison scores depend only on the probe and reference images (no persearch normalization). If our assumptions are valid, false rejects will hardly ever happen. 單次執行,且比對分數僅取決於探測影像和參考影像(無逐次搜尋正規化)。如果我們的假設有效,誤拒將幾乎不會發生。
Be aware that the performance of a face recognition system also depends on other factors than distance, like, e. g., illumination, pose, exposure, ageing, which can impact the mate score distribution. 請注意,人臉辨識系統的效能也取決於距離以外的其他因素,例如照明、姿勢、曝光、老化等,這些因素都可能影響配對分數的分佈。
Demonstration: Let U_(n)U_{n} be a random variable representing a non-mated comparison score. Its mean is mu_(n)\mu_{n} and its standard deviation is sigma_(n)\sigma_{n}. Assume that the probability distribution of U_(n)U_{n} is a normal distribution f_(n)f_{n}. Let U_(m)U_{m} be a random variable representing a mated comparison score. Its mean is mu_(m)\mu_{m} and its standard deviation is sigma_(m)\sigma_{m}. Assume that the probability distribution of U_(m)U_{m} is a normal distribution f_(m)f_{m}. Furthermore, assume that the probability distribution of U_(n)U_{n} and the probability distribution of U_(m)U_{m} are independent. 示範:令 U_(n)U_{n} 為代表非配對比對分數的隨機變數。其平均值為 mu_(n)\mu_{n} ,標準差為 sigma_(n)\sigma_{n} 。假設 U_(n)U_{n} 的機率分佈為常態分佈 f_(n)f_{n} 。令 U_(m)U_{m} 為代表配對比對分數的隨機變數。其平均值為 mu_(m)\mu_{m} ,標準差為 sigma_(m)\sigma_{m} 。假設 U_(m)U_{m} 的機率分佈為常態分佈 f_(m)f_{m} 。此外,假設 U_(n)U_{n} 的機率分佈與 U_(m)U_{m} 的機率分佈是獨立的。
Then, X=U_(m)-U_(n)X=U_{m}-U_{n} is also a random variable with a normal distribution f_(X)f_{X}. Its mean is mu_(X)\mu_{X} and its standard deviation is sigma_(X)\sigma_{X}. The mean and the standard deviation of XX can also be determined: 則 X=U_(m)-U_(n)X=U_{m}-U_{n} 也是一個具有常態分佈 f_(X)f_{X} 的隨機變數。其平均值為 mu_(X)\mu_{X} ,標準差為 sigma_(X)\sigma_{X} 。 XX 的平均值和標準差也可以確定:
Figure E. 31 shows examples of mated and non-mated score distributions and the distribution of differences of mated and non-mated scores. 圖 E.31 顯示了配對和非配對分數分佈以及配對和非配對分數差異分佈的範例。
In this study, d^(')in[12,7;14,8]d^{\prime} \in[12,7 ; 14,8]. So, d^(')=(|mu_(m)-mu_(n)|)/(sqrt(sigma_(m)^(2)+sigma_(n)^(2)))=(|mu_(X)|)/(sigma_(X)) > 12d^{\prime}=\frac{\left|\mu_{m}-\mu_{n}\right|}{\sqrt{\sigma_{m}{ }^{2}+\sigma_{n}{ }^{2}}}=\frac{\left|\mu_{X}\right|}{\sigma_{X}}>12. So, |mu_(X)| > 12*sigma_(X)\left|\mu_{X}\right|>12 \cdot \sigma_{X}. 在本研究中, d^(')in[12,7;14,8]d^{\prime} \in[12,7 ; 14,8] 。因此, d^(')=(|mu_(m)-mu_(n)|)/(sqrt(sigma_(m)^(2)+sigma_(n)^(2)))=(|mu_(X)|)/(sigma_(X)) > 12d^{\prime}=\frac{\left|\mu_{m}-\mu_{n}\right|}{\sqrt{\sigma_{m}{ }^{2}+\sigma_{n}{ }^{2}}}=\frac{\left|\mu_{X}\right|}{\sigma_{X}}>12 。因此, |mu_(X)| > 12*sigma_(X)\left|\mu_{X}\right|>12 \cdot \sigma_{X} 。
Figure E. 31 - Distribution of differences of mated and non-mated scores 圖 E. 31 - 配對與非配對分數差異的分佈
Let XX be normally distributed with mean |mu_(X)|=12*sigma_(X)\left|\mu_{X}\right|=12 \cdot \sigma_{X} and standard deviation sigma_(X)\sigma_{X}. 令 XX 為平均值 |mu_(X)|=12*sigma_(X)\left|\mu_{X}\right|=12 \cdot \sigma_{X} 和標準差 sigma_(X)\sigma_{X} 的常態分佈。
Let X^(')=(X-12*sigma_(X))/(sigma_(X))X^{\prime}=\frac{X-12 \cdot \sigma_{X}}{\sigma_{X}}. Then, X^(')X^{\prime} is normally distributed with mean 0 and standard deviation 1 , and P(X^(') <= -12)=2,15*10^(-32)P\left(X^{\prime} \leq-12\right)=2,15 \cdot 10^{-32}. 令 X^(')=(X-12*sigma_(X))/(sigma_(X))X^{\prime}=\frac{X-12 \cdot \sigma_{X}}{\sigma_{X}} 。則 X^(')X^{\prime} 為平均值 0 和標準差 1 的常態分佈,且 P(X^(') <= -12)=2,15*10^(-32)P\left(X^{\prime} \leq-12\right)=2,15 \cdot 10^{-32} 。
Based on the data collected for this study, camera subject distance does not have a great influence on face verification performance over the range of camera subject distances investigated. Over a database of excellent-quality face images of 435 test subjects, captured at single capture sessions per test subject from different camera subject distances, several face verification algorithms avoided verification errors altogether. 根據本研究收集的資料,在所調查的相機主體距離範圍內,相機主體距離對人臉驗證效能沒有太大影響。在一個包含 435 名受試者、於單次拍攝會話中從不同相機主體距離拍攝的優質人臉影像資料庫中,有數個人臉驗證演算法完全避免了驗證錯誤。
It is noticeable that the average d^(')d^{\prime} value is not sensitive at all to a reference distance above 0,7m0,7 \mathrm{~m}, i.e., below 7,1%7,1 \% of magnification distortion. Nearly symmetrically, the average d^(')d^{\prime} value is not sensitive at all to a probe distance above 0,7m0,7 \mathrm{~m}, i.e., below 7,1%7,1 \% magnification distortion. Even at 0,5m0,5 \mathrm{~m}, i.e., at 10%10 \% magnification distortion d^(')d^{\prime}, decreases only 15%15 \% to 12,7 versus the maximum value of 14,8 obtained with magnification distortion below 7,1%7,1 \%. 值得注意的是,平均 d^(')d^{\prime} 值對於高於 0,7m0,7 \mathrm{~m} 的參考距離完全不敏感,亦即,低於 7,1%7,1 \% 的放大失真。幾乎對稱地,平均 d^(')d^{\prime} 值對於高於 0,7m0,7 \mathrm{~m} 的探測距離完全不敏感,亦即,低於 7,1%7,1 \% 的放大失真。即使在 0,5m0,5 \mathrm{~m} ,亦即,在 10%10 \% 放大失真 d^(')d^{\prime} ,相較於在放大失真低於 7,1%7,1 \% 時獲得的最大值 14.8,僅減少 15%15 \% 至 12.7。
The enrolment and verification images should be captured from a similar distance whenever possible. 註冊和驗證影像應盡可能從相似的距離擷取。
E. 7 Example of exposure metering at various spots on a subject E. 7 主體不同位置的曝光測量範例
The exposure value (EV) is the value given to any combination of shutter speed and aperture (F-stop) that results in the same exposure. By definition, an EV value of 0 corresponds to a shutter speed of 1 second and an aperture of F1,0, for a film speed or equivalent image sensor sensitivity of ISO 100. EV is defined by: 曝光值 (EV) 是指快門速度和光圈 (F 值) 任何組合所產生的相同曝光值。根據定義,EV 值為 0 對應於快門速度 1 秒和光圈 F1.0,適用於 ISO 100 的底片速度或等效影像感測器靈敏度。EV 的定義為:
where 其中 FF is the F-stop setting; FF 是光圈設定; T quadT \quad is the exposure time.
T quadT \quad
為曝光時間。
A change of 1 EV corresponds to a one F-stop aperture increase or decrease or a halving or doubling of the exposure time. 1 EV 的變化對應於光圈增加或減少一個 F-stop,或曝光時間減半或加倍。
Bibliography 參考書目
[1] AAMVA DL/Identifier-2000, American Association of Motor Vehicle Administrators National Standard for the Driver License/Identification Card [1] AAMVA DL/Identifier-2000,美國汽車監理人員協會駕駛執照/身分證國家標準
[2] Anthropometry of the Head and Face, second edition, Leslie G. Farkas, Raven Press, New York, 1994 [2] 《頭部與臉部的人體測量學》,第二版,Leslie G. Farkas,Raven Press,紐約,1994 年
[3] C-Cube Microsystems, JPEG File Interchange Format (JFIF), Version 1.02 [3] C-Cube Microsystems,JPEG 檔案交換格式 (JFIF),版本 1.02
[4] International Color Consortium, Available at www.color.org [4] 國際色彩聯盟,可於 www.color.org 取得
[5] NIST Best Practice Recommendation For The Capture Of Mugshots, Version 2.0, 1997 [5] NIST 拍攝臉部照片的最佳實務建議,2.0 版,1997 年
[6] PIMA 7667:2001, Photography - Electronic Still Picture Imaging - Extended sRGB Color Encoding - e-sRGB [6] PIMA 7667:2001,攝影 — 電子靜態影像 — 擴展 sRGB 色彩編碼 — e-sRGB
[7] The Methods of Plane Projective Geometry Based on the Use of General Homogenous Coordinates, E.A. Maxwell, Cambridge University Press, 1960 [7] 平面投影幾何學方法,基於一般齊次座標的使用,E.A. Maxwell,劍橋大學出版社,1960 年
[8] Verständigung über ein gemeinsames craniometrisches Verfahren (Frankfurter Verständigung). Ranke, J. (ed.). (1884). Archive Anthropologie, 15, 1-8 [8] Verständigung über ein gemeinsames craniometrisches Verfahren (Frankfurter Verständigung)。Ranke, J. (ed.)。(1884)。Archive Anthropologie,15,1-8
[9] ISO 20473:2007, Optics and photonics — Spectral bands [9] ISO 20473:2007,光學與光子學 — 光譜帶
[10] ISO/IEC 8825-1:2015, Information technology - ASN. 1 encoding rules: Specification of Basic Encoding Rules (BER), Canonical Encoding Rules (CER) and Distinguished Encoding Rules (DER) - Part 1 [10] ISO/IEC 8825-1:2015,資訊科技 — ASN.1 編碼規則:基本編碼規則 (BER)、規範編碼規則 (CER) 與區別編碼規則 (DER) 規範 — 第 1 部
[11] IEC 61966-8:2001, Multimedia systems and equipment - Colour measurement and management Part 8: Multimedia colour scanners [11] IEC 61966-8:2001,多媒體系統與設備 — 色彩量測與管理 第 8 部分:多媒體色彩掃描器
[12] ANSI IT8.7/2-1993 (R2013) Graphic technology - Color reflection target for input scanner calibration [12] ANSI IT8.7/2-1993 (R2013),圖形技術 — 輸入掃描器校正用色彩反射目標
[13] ISO 12641-1, Graphic technology - Prepress digital data exchange - Colour targets for input scanner calibration - Part 1: Colour targets for input scanner calibration [13] ISO 12641-1,圖形技術 — 印前數位資料交換 — 輸入掃描器校正用色彩目標 — 第 1 部分:輸入掃描器校正用色彩目標
[14] ISO/CIE 11664-3, Colorimetry - Part 3: CIE tristimulus values [14] ISO/CIE 11664-3,色彩測量學 — 第 3 部分:CIE 三刺激值
[15] ISO/IEC 11664-4, Colorimetry - Part 4: CIE 1976 L ^(**)a^(**)b^(**)^{*} a^{*} b^{*} colour space [15] ISO/IEC 11664-4,色彩測量學 — 第 4 部分:CIE 1976 L ^(**)a^(**)b^(**)^{*} a^{*} b^{*} 色彩空間
[16] IEC 61966-2-1:1999, Multimedia systems and equipment - Colour measurement and management - Part 2-1: Colour management - Default RGB colour space - sRGB [16] IEC 61966-2-1:1999,多媒體系統與設備 — 色彩測量與管理 — 第 2-1 部分:色彩管理 — 預設 RGB 色彩空間 — sRGB
[17] Rochester Institute of Technology https://www.cs.rit.edu/~ncs/color/t_convert.html#RGB [17] 羅徹斯特理工學院 https://www.cs.rit.edu/~ncs/color/t_convert.html#RGB
[18] Report Field Application, Materials Characterization: UV/Vis/NIR Spectroscopy, Jillian F. Dlugos, Jeffrey L. Taylor, Ph.D., PerkinElmer, Inc., 2012 [18] 應用領域報告,材料特性:UV/Vis/NIR 光譜學,Jillian F. Dlugos,Jeffrey L. Taylor 博士,PerkinElmer, Inc.,2012 年
[19] Digital imaging & image processing techniques for the comparison of human hair features, Carolyn Jane McLaren, University of Canberra, Faculty of Applied Science, 2012
[20] Grading of Iris Color with an Extended Photographic Reference Set, Luuk Franssen, Joris E. Coppens, Thomas J.T.P. van den Berg , , Netherlands Institute for Neuroscience, an institute of the Royal, Netherlands Academy of Arts and Sciences. Amsterdam, The Netherlands, 2008 [20] Grading of Iris Color with an Extended Photographic Reference Set, Luuk Franssen, Joris E. Coppens, Thomas J.T.P. van den Berg, Netherlands Institute for Neuroscience, an institute of the Royal, Netherlands Academy of Arts and Sciences. Amsterdam, The Netherlands, 2008
[21] IEC 62676-4, Video surveillance systems for use in security applications - Part 4: Application guidelines
[22] Spectral Filter Optimization for the Recovery of Parameters Which Describe Human Skin, Stephen J. Preece and Ela Claridge, IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 26, No. 7, July 2004
[23] CIE Delta E, 2000. Improvement to industrial color-difference evaluation. Vienna: CIE Publication No. 142-2001, Central Bureau of the CIE; 2001
[24] ISO 16067-1:2003, Photography - Spatial resolution measurements of electronic scanners for photographic images - Part 1: Scanners for reflective media
[25] Characterising the variations in ethnic skin colours: a new calibrated data base for human skin. K. Xiao, J. M. Yates, F. Zardawi, S. Sueeprasan, N. Liao, L. Gill, C.Li and S. Wuerger. Skin Research and Technology 2017; 23: 21-29. Published by John Wiley & Sons Ltd, 2016
[26] Measuring H.S.C., Mengmeng Wang, Kaida Xiao, Sophie Wuerger, Vien Cheung, Ming Ronnier Luo. 23rd Color and Imaging Conference Final Program and Proceedings, Society for Imaging Science and Technology, 2015
[27] ISO/IEC 29794-1, Information technology - Biometric sample quality - Part 1: Framework [27] ISO/IEC 29794-1,資訊科技—生物辨識樣本品質—第 1 部分:框架
[28] ISO/IEC 19795-1, Information technology - Biometric performance testing and reporting Part 1: Principles and framework [28] ISO/IEC 19795-1,資訊科技—生物辨識效能測試與報告—第 1 部分:原則與框架
[29] Bolle R.M., Pankanti S., Ratha N.K., Evaluation techniques for biometrics-based authentication systems (FRR). In Proceedings of the 15th International Conference on Pattern Recognition ICPR, volume 2, 2000 [29] Bolle R.M.、Pankanti S.、Ratha N.K.,基於生物辨識的驗證系統評估技術 (FRR)。收錄於第 15 屆國際模式識別會議 (ICPR) 論文集,第 2 卷,2000 年
[30] Technical Guideline TR-03121-3: Biometrics for public sector applications, Part 3: Application Profiles and Function Modules, Volume 1: Verification scenarios for ePassport and Identity Card, Version 3.0.1. 2013 [30] 技術指南 TR-03121-3:公共部門應用之生物辨識,第 3 部分:應用設定檔與功能模組,第 1 卷:電子護照與身分證之驗證情境,版本 3.0.1。2013 年
[31] ISO/IEC 19785 (all parts), Information technology - Common Biometric Exchange Formats Framework [31] ISO/IEC 19785(所有部分),資訊科技—通用生物識別交換格式框架
[32] ISO/IEC 14496-1, Information technology - Coding of audio-visual objects - Part 1: Systems [32] ISO/IEC 14496-1,資訊科技—視聽物件編碼—第 1 部分:系統
[33] Netpbm image format, http://netpbm.sourceforge.net/doc/pgm.html [33] Netpbm 影像格式,http://netpbm.sourceforge.net/doc/pgm.html
[34] Netpbm color image format, http://netpbm.sourceforge.net/doc/ppm.html [34] Netpbm 彩色影像格式,http://netpbm.sourceforge.net/doc/ppm.html
[35] W3C XML sources: http://www.W3C.org/XML/SCHEMA.html#Tools [35] W3C XML 來源:http://www.W3C.org/XML/SCHEMA.html#Tools
[36] ITU-T ASN. 1 sources: https://www.itu.int/en/ITU-T/asn1/Pages/Tools.aspx [36] ITU-T ASN.1 來源:https://www.itu.int/en/ITU-T/asn1/Pages/Tools.aspx
[37] ISO 15739:2017, Photography - Electronic still-picture imaging - Noise measurements [37] ISO 15739:2017,攝影 — 電子靜態影像 — 雜訊量測
[38] ICAO Technical Report, Portrait Quality (Reference Facial Images for MRTD): https://www.icao .int/Security/FAL/TRIP/Documents/TR%20-%20Portrait%20Quality%20v1.0.pdf [38] ICAO 技術報告,人像品質(機器可讀旅行證件的參考臉部影像):https://www.icao.int/Security/FAL/TRIP/Documents/TR%20-%20Portrait%20Quality%20v1.0.pdf
This page has been left intentionally blank. 此頁面刻意留白。
导出提示
謹慎下載超過 300 頁以上 PDF,容易記憶體溢出。
PDF Pro 下載需借用瀏覽器的打印功能,出現打印界面後,在【打印機】處選擇【另存為 PDF】,最後點擊【保存】即可。