1、INTERNATIONAL STANDARD ISO/IEC 10918-2 First edition 1995-08-I 5 Information technology - Digital compression and coding of continuous-tone still images: Compliance testing Technologies de /information - Compression numkique et codage des images fixes de nature photographique: Test de conformit Refe
2、rence number ISO/IEC 10918-2:1995(E) Adopted by INCITS (InterNational Committee for Information Technology Standards) as an American National Standard.Date of ANSI Approval: 1/11/99Published by American National Standards Institute,25 West 43rd Street, New York, New York 10036Copyright 2002 by Infor
3、mation Technology Industry Council (ITI).All rights reserved.These materials are subject to copyright claims of International Standardization Organization (ISO), InternationalElectrotechnical Commission (IEC), American National Standards Institute (ANSI), and Information Technology Industry Council(
4、ITI). Not for resale. No part of this publication may be reproduced in any form, including an electronic retrieval system, withoutthe prior written permission of ITI. All requests pertaining to this standard should be submitted to ITI, 1250 Eye Street NW,Washington, DC 20005.Printed in the United St
5、ates of AmericaISOAEC 10918-2:1995(E) CONTENTS 1 Scope 2 Normative references . 3 Definitions, abbreviations, symbols, and conventions . 4 General . 5 Compressed data format compliance testing 6 Encoder compliance tests . 7 Decoder compliance tests . Annex A - Procedures for determining generic enco
6、der and decoder compliance . Annex B - Quantization tables for generic compliance testing of DCT-based processes . Annex C - Compressed test data stream structure for generic decoder compliance tests . Annex D - Construction of application-specific compliance tests Annex E - Compliance test data for
7、 testing of greater computational accuracy . Annex F - Specification of supported parameter ranges . Annex H - Examples and guidelines . Page 1 1 1 4 6 19 20 25 28 30 51 53 55 58 Q ISO/IEC 1995 All rights reserved. Unless otherwise specified, no part of this publication may be reproduced or utilized
8、 in any form or by any means, electronic or mechanical, including photocopying and microfilm, without permission in writing from the publisher. ISO/IEC Copyright Office l Case postale 56 l CH-1211 Gedve 20 0 Switzerland Printed in Switzerland ii 0 ISO/IEc ISO/IEC 1091%2:1995(E) Foreword IS0 (the Int
9、ernational Organization for Standardization) and IEC (the Inter- national Electrotechnical Commission) form the specialized system for worldwide standardization. National bodies that are members of IS0 or IEC participate in the development of International Standards through technical committees esta
10、blished by the respective organization to deal with particular fields of technical activity. IS0 and IEC technical committees collaborate in fields of mutual interest. Other international organizations, governmental and non-governmental, in liaison with IS0 and IEC, also take part in the work. In th
11、e field of information technology, IS0 and IEC have established a joint technical committee ISO/IEC JTC 1. Draft International Standards adopted by the joint technical committee are circulated to national bodies for voting. Publication as an International Standard requires approval by at least 75 %
12、of the national bodies casting a vote. International Standard ISO/IEC 10918-2 was prepared by Joint Technical Com- mittee ISO/IEC JTC 1, Information technology, Subcommittee SC 29, Coding of audio, picture, multimedia and hypermedia information, in collaboration with ITU-T. The identical text is pub
13、lished as ITU-T Recommendation T.83. ISO/IEC 10918 consists of the following parts, under the general title Information technology - Digital compression and coding of continuous-tone still images: - Part 1: Requirements and guidelines - Part 2: Compliance testing Annexes A to D form an integral part
14、 of this part of ISO/IEC 10918. Annexes E to H are for information only. . . . 111 ISO/IEC 1091 - specifies compliance tests for the ITU-T Rec. T.8 1 I ISO/IEC 109 18- 1 encoding processes; - specifies compliance tests for the ITU-T Rec. T.8 1 I ISO/IEC 10918-l decoding processes; - specifies a meth
15、od for constructing application-specific compliance tests; - gives guidance and examples on how to implement these tests in practice. This Specification specifies normative generic compliance tests for the ITU-T Rec. T.81 I ISO/IEC 10918-l encoding and decoding processes. These compliance tests are
16、applicable to “stand-alone” generic implementations of one or more of the encoding and decoding processes specified in ITU-T Rec. T.81 I ISO/IEC 10918-l. Among the purposes of these tests is to ensure that generic encoder (and decoder) implementations compute the discrete cosine transform (DCT) and
17、quantization functions with sufficient accuracy. 2 Normative references The following ITU-T Recommendations and International Standards contain provisions which, through reference in this text, constitute provisions of this Recommendation I International Standard. At the time of publication, the edi
18、tions indicated were valid. All Recommendations and Standards are subject to revision, and parties to agreements based on this Recommendation I International Standard are encouraged to investigate the possibility of applying the most recent editions of the Recommendations and Standards indicated bel
19、ow. Members of IEC and IS0 maintain registers of currently valid International Standards. The Telecommunication Standardization Bureau of the ITU-T maintains a list of the currently valid ITU-T Recommendations. 2.1 Additional references - IS0 5807:1985, Information processing - Documentation symbols
20、 and conventions for data, program and system flowcharts, program network charts and system resources charts. 3 Definitions, abbreviations, symbols, and conventions 3.1 Definitions For the purposes of this Specification, the following definitions apply. 3.1.1 (coding) process 1: Coding process with
21、baseline sequential DCT, g-bit sample precision. 3.1.2 (coding) process 2: Coding process with extended sequential DCT, Huffman coding, g-bit sample precision. 3.1.3 (coding) process 3: Coding process with extended sequential DCT, arithmetic coding, g-bit sample precision. 3.1.4 (coding) process 4:
22、Coding process with extended sequential DCT, Huffman coding, 12-bit sample precision. ITU-T Rec. T.83 (1994 E) ISO/IEC 10918-2 : 1995 (E) 3.1.5 (coding) process 5: Coding process with extended sequential DCT, arithmetic coding, 12-bit sample precision. 3.1.6 (coding) process 6: Coding process with s
23、pectral selection only, Huffman coding, 8-bit sample precision. 3.1.7 (coding) process 7: Coding process with spectral selection only, arithmetic coding, g-bit sample precision. 3.1.8 (coding) process 8: Coding process with spectral selection only, Huffman coding, 12-bit sample precision. 3.1.9 (cod
24、ing) process 9: Coding process with spectral selection only, arithmetic coding, 12-bit sample precision. 3.1.10 (coding) process 10: Coding process with full progression, Huffman coding, 8-bit sample precision. 3.1.11 (coding) process 11: Coding process with full progression, arithmetic coding, 8-bi
25、t sample precision. 3.1.12 (coding) process 12: Coding process with full progression, Huffman coding, 1Zbit sample precision. 3.1.13 (coding) process 13: Coding process with full progression, arithmetic coding, 1Zbit sample precision. 3.1.14 (coding) process 14: Coding process with lossless, Huffman
26、 coding, 2- through 16-bit sample precision. 3.1.15 (coding) process 15: Coding process with lossless, arithmetic coding, 2- through 16-bit sample precision. 3.1.16 (coding) process 16: Coding process with extended sequential DCT, Huffman coding, g-bit sample precision in hierarchical mode. 3.1.17 (
27、coding) process 17: Coding process with extended sequential DCT, arithmetic coding, each component is a two dimensional array of 8 x 8 blocks stored left-to-right, top-to-bottom order; each 8 x 8 block has 64 coefficients stored in zigzag order; and each coefficient is represented by two bytes, the
28、most significant byte first. This data includes the blocks which are padded to complete an MCU on the right and bottom of the image. (Distributed as part of the compliance test data.) 3.1.35 encoder reference test data: Quantized DCT coefficient data generated by the reference FDCT and reference qua
29、ntizer from the source image test data to be used in the DCT-based encoder compliance tests. (Distributed as part of the compliance test data.). 3.1.36 generic: Applicable to a broad range of applications, i.e. application independent. 3.1.37 orthogonal representation: The 2-dimensional row-column f
30、ormat illustrated in Figure A.5 in ITU-T Rec. T.81 I ISO/IEC 10918-l. 3.1.38 quantized coefficient validation data: Quantized DCT coefficient data generated from the source image validation test data to be used in the DCT-based encoder validation tests. (Distributed as part of the compliance test da
31、ta.). 3.1.39 reference DCT-based decoder: An embodiment of the DCT-based decoding processes which generates the decoder reference test data. It consists of an entropy decoder, a dequantizer, and the reference IDCT. 3.1.40 reference DCT-based encoder: An embodiment of the DCT-based encoding processes
32、 which generated the DCT-based compressed image test data streams. It consists of the reference FDCT, the reference quantizer, and an entropy encoder. 3.1.41 reference forward discrete cosine transform; reference FDCT: A double precision (64-bit) floating point embodiment of the FDCT described in A.
33、3.3 of ITU-T Rec. T.81 I ISO/IEC 10918-l. 3.1.42 reference inverse discrete cosine transform; reference IDCT: A double precision (64-bit) floating point embodiment of the IDCT described in A.3.3 of ITU-T Rec. T.81 I ISO/IEC 10918-l. 3.1.43 reference quantizer: An embodiment of the quantization descr
34、ibed in A.3.4 in ITU-T Rec. T.81 I ISO/IEC 10918-l. 3.1.44 source image test data: The data sets to be used as input to the encoder compliance tests. This data is a sequence of pseudo-random numbers generated with uniform distribution over the range from 0 to 255. The algorithm used to generate this
35、 data is described in Annex A of CCITT Recommendation H.26 1. (This data is distributed as part of the compliance test data.). 3.1.45 table specification test data (stream): Table specification data generated to test decoder compliance with abbreviated format compressed data. (Distributed as part of
36、 the compliance test data.) 3.2 Abbreviations The abbreviations used in this Specification are listed below. 3.2.1 arith.: An abbreviation for arithmetic coding. 3.2.2 Huff.: An abbreviation for Huffman coding. 3.3 Symbols The symbols used in this Specification are listed below. 3.3.1 Bi: quantizati
37、on value at the ith row and jth column in the quantization tables defined in Annex B, appears in Annex E. 3.3.2 DF: differential frame flag, appears in flow charts in clause 5. ITU-T Rec. T.83 (1994 E) 3 ISO/IEC 10918-2 : 1995 (E) 3.3.3 Eu: quantization value at the ith row and jth column in the qua
38、ntization tables used in testing for greater accuracy defined in Annex E. 3.3.4 F: the scale factor used to generate Eij from Bij as defined in E. 1. 3.3.5 FS: first scan in frame flag, appears in flow charts in clause 5. 3.3.6 G: guaranteed in compressed data, appears in Tables 1 to 5 in clause 5.
39、3.3.7 H-L: hierarchical lossless processes, appears in Table G. 1. 3.3.8 H-S: hierarchical sequential DCT-based processes without final lossless scans, appears in Table G. 1. 3.3.9 HP: hierarchical progression flag, appears in flow charts in clause 5. 3.3.10 LL: lossless processes, appears in Table
40、G.1. 3.3.11 o: optional in compressed data, appears in tables in clause 5. 3.3.12 P(FULL): full progressive DCT-based processes with both spectral selection and successive approximation, appears in Table G. 1. 3.3.13 P(SA): progressive DCT-based successive approximation processes, appears in Table G
41、. 1. 3.3.14 P(SS): progressive DCT-based spectral selection processes, appears in Table G. 1. 3.3.15 RI: restart interval flag, appears in flow charts in clause 5. 3.3.16 S(B): baseline sequential DCT-based process, appears in Table G. 1. 3.3.17 S(E): extended sequential DCT-based processes, appears
42、 in Table G. 1. 3.4 Conventions The flowcharts use the conventions given in IS0 5807. One of the conventions is that arrows are not needed when the flow is from left-to-right and from top-to-bottom. Arrows are sometimes used in such cases to increase clarity. 4 General The purpose of this clause is
43、to give an informative overview of this Specification and the principles underlying it. Another purpose is to introduce some of the terms which are defined in clause 3. (Terms defined in clause 3 of ITU-T Rec. T.81 I ISO/IEC 10918-I continue to apply in this Specification.) ITU-T Rec. T.83 I ISO/IEC
44、 10918-2 concerns compliance testing for embodiments of the elements specified in ITU-T Rec. T.8 1 I ISO/IEC 10918-I. For encoders and decoders - embodiments of the ITU-T Rec. T.8 1 I ISO/IEC 109 18-l encoding and decoding processes - this document makes a distinction between GENERIC embodiments and
45、 APPLICATION-SPECIFIC embodiments. For the former, compliance tests themselves are specified herein; for the latter, this document specifies a method for defining compliance tests. Compliance tests are also specified for compressed data streams - embodiments of the ITU-T Rec. T.81 I ISO/IEC 10918-l
46、compressed data formats. NOTE - Like many compliance tests, those described in this Specification for generic encoders and decoders are not exhaustive tests of their respective functional specifications. Therefore, passing these tests does not guarantee complete functional correctness. This observat
47、ion has two implications: 1) the tests do not fully guarantee complete interoperability between independently-implemented encoders and decoders; and 2) the tests for embodiments of the DCT-based processes do not guarantee that encoders or decoders will have some well defined image-quality-producing
48、capability. These limitations are discussed in more detail below. 4.1 Purpose of the compliance tests The purpose of compliance tests is to provide designers, manufacturers, or users of a product with a set of procedures for determining whether the product meets a specified set of requirements with
49、some confidence. In addition, the compliance tests specified herein are intended to achieve the following specific goals: - increase the likelihood of compressed data interchange; - decrease the likelihood that DCT-based encoders or decoders will yield reduced image quality as a result of computing the DCT or quantization procedures with insufficient accuracy: - help implementors to meet the ITU-T Rec. T.81 I ISO/IEC 10918-l requirements for encoders and decoders as fully as possible. 4 ITU-T Rec. T.83 (1994 E) ISO/IEC 10918-2 : 1995 (E) 4.2 Compressed data compliance tests The aim of the com
