ITU-T T 81-1992 Information Technology - Digital Compression and Coding of Continuous-Tone Still Images - Requirements and Guidelines - Terminal Equipment and Protocols for Telemat.pdf

1、INTERNATIONAL TELECOMMUNICATION UNION CCITT THE INTERNATIONAL TELEGRAPH AND TELEPHONE CONSULTATIVE COMMITTEE TERMINAL EQUIPMENT AND PROTOCOLS FOR TELEMATIC SERVICES T.81 (09/92) INFORMATION TECHNOLOGY - DIGITAL COMPRESSION AND CODING REQUIREMENTS AND GUIDELINES OF CONTINUOUS-TONE STILL IMAGES - I Re

2、commendation T.81 ITU-T RECMN*T-83 93 4862593 0586637 7T9 Foreword ITU (International Telecommunication Union) is the United Nations Specialized Agency in the field of telecommunications. The CCITT (the International Telegraph and Telephone Consultative Committee) is a permanent organ of the ITU. So

3、me 166 member countries, 68 telecom operating entities, 163 scientific and industrial organizations and 39 international organizations participate in CCIT which is the body which sets world telecommunications standards (Recommendations). The approval of Recommendations by the members of CCITT is cov

4、ered by the procedure laid down in CCIT Resolution No. 2 (Melbourne, 1988). In addition, the Plenary Assembly of CCITT, which meets every four years, approves Recommendations submitted to it and establishes the study programme for the following period. In some areas of information technology, which

5、fall within CCITTs purview, the necessary standards are prepared on a collaborative basis with IS0 and IEC. The text of CCIT Recommendation T.81 was approved on 18th September 1992. The identical text is also published as ISO/IEC International Standard 109 18-1. CC NOTE In this Recommendation, the e

6、xpression “Administration” is used for conciseness to indicate both a telecommunication administration and a recognized private operating agency. O ITU 1993 Ail rights reserved. No part of this publication may be reproduced or utilized in any form or by any means, electronic or mechanical, including

7、 photocopying and microfilm, without permission in writing from the ITU. Conens Inuoduction 1 scope . 2 Normative references . 3 Definitions. abbreviations and symbols 4 General 5 Interchange format requirements . 6 Encoder requirements 7 Decoder requirements Annex A - Mathematical definitions Annex

8、 B - Compressed data formats Annex C - Huffman table specification . Annex D - Arithmetic coding . Annex E - Encoder and decoder control procedures . Annex F - Sequential DCT-based mode of operation Annex G - Progressive DCT-based mode of operation Annex H - Lossless mode of operation . Annex J - Hi

9、erarchical mode of operation I Annex K - Examples and guidelines Annex L - Patents Annex M - Bibliography CCITT Rec . T.81 (1992 E) Page . 111 1 12 23 23 23 24 31 50 54 77 87 119 132 137 143 179 181 i ITU-T RECMN*T*8L 93 M 486259% 058bb21 357 Introduction This CCITT Recommendation I ISOAEC Internati

10、onal Standard was prepared by CCITT Study Group VI11 and the Joint Photographic Experts Group (JPEG) of ISO/IEC JTC l/SC 29WG 10, This Experts Group was formed in 1986 to establish a standard for the sequential progressive encoding of continuous tone grayscale and colour images. Digitul Compression

11、and Coding of Continuous-tone Still images, is published in two parts: - Requirements and guidelines: - Compliance testing. This part, Part 1, sets out requirements and implementation guidelines for continuous-tone still image encoding and decoding processes, and for the coded representation of comp

12、ressed image data for interchange between applications. These processes and representations are intended to be generic, that is, to be applicable to a broad range of applications for colour and grayscale still images within communications and computer systems. Part 2, sets out tests for determining

13、whether implementations comply with the requirments for the various encoding and decoding processes specified in Part 1. The users attention is called to the possibility that - for some of the coding processes specified herein - compliance with this Recommendation I International Standard may requir

14、e use of an invention covered by patent nghts. See Annex L for further information. The requirements which these processes must satisfy to be useful for specific image communications applications such as facsimile, Videotex and audiographic conferencing are defined in CCIT Recommendation T.80. The i

15、ntent is that the generic processes of Recommendation T.80 will be incorporated into the varioiis CCITT Recommendations for terminal equipment for these applications. In addition to the applications addressed by the CCITT and ISODEC, the JPEG committee has developped a compression standard to meet t

16、he needs of other applications as weil, including desktop publishing, graphic arts, medical imaging and scientific imaging. Annexes A, B, C, D, E, F, G, H and J are normative, and thus form an integral part of this Specification. Annexes K, L and M are informative and thus do not form an integral pa

17、rt of this Specification. This Specification aims to follow the guidelines of CCITT and ISO/IEC JTC 1 on Rules for presentation of CC17T I ISO/IEC common text. CCITT Rec. T.81 (1992 E) iii - ITU-T RECMN*T=BL 93 Y8b259L 0586622 293 ISO/IEC 10918-1 1993(E) INTERNATIONAL STANDARD 1 CCITT RECOMMENDATION

18、 INFORMATION TECHNOLOGY - DIGITAL COMPRESSION AND CODING OF CONTINUOUS-TONE STILL IMAGES - REQUIREMENTS AND GUIDELINES 1 Scope This CCITT Recommendation I International Standard is applicable to continuous-tone - grayscale or colour - digital still image data. It is applicable to a wide range of app

19、lications which require use of compressed images. It is not applicable to bi-level image data. This Specification - - - - NOTE - This Specification does not specis a complete coded image representation. Such representations may include certain parameters, such as aspect ratio, component sample regis

20、tration, and colour space designation, which are application- dependen t. specifies processes for converting source image data to compressed image data; specifies processes for converting compressed image data to reconstructed image data: gives guidance on how to implement these processes in practic

21、e; specifies coded representations for compressed image data. 2 Normative references The following CCITT Recommendations and International Standards contain provisions which, through reference in this text, constitute provisions of this CCITT Recommendation I International Standard. At the time of p

22、ublication, the editions indicated were valid. All Recommendations and Standards are subject to revision, and parties to agreements based on this CCIT Recommendation I International Standard are encouraged to investigate the possibility of applying the most recent edition of the Recommendations and

23、Standards listed below. Members of EC and IS0 maintain registers of currently valid International Standards. The CCITT Secretariat maintains a list of currently valid CCITT Recommendations. - CCIV Recommendation T.80 (1992), Common components for image compression and communication - Basic principle

24、s. 3 Definitions, abbreviations and symbols 3.1 Definitions and abbreviations For the purposes of this Specification, the following definitions apply. 3.1.1 abbreviated format: A representation of compressed image data which is missing some or all of the table specifications required for decoding, o

25、r a representation of table-specification data without frame headers, scan headers, and entropy-coded segments. 3.1.2 3.1.3 symbols from the sequence of bits produced by the arithmetic encoder. 3.1.4 recursive subdivision of the probability of the sequence of symbols coded up to that point. 3.1.5 es

26、tablished for a particular application. AC coefficient: Any DCT coefficient for which the frequency is not zero in at least one dimension. (adaptive) (binary) arithmetic decoding: An entropy decoding procedure which recovers the sequence of (adaptive) (binary) arithmetic encoding: An entropy encodin

27、g procedure which codes by means of a application environment: The standards for data representation, communication, or storage which have been CCITT Rec. T.81(1992 E) 1 ITU-T RECMN*T.BL 73 W 4862591 0586623 L2T W ISO/IEC 10918-1 : 1993(E) 3.1.6 arithmetic decoder: An embodiment of arithmetic decodi

28、ng procedure. 3.1.7 3.1.8 Specification, and which is required for all DCT-based decoding processes. 3.1.9 3.1.10 3.1.11 arithmetic encoder: An embodiment of arithmetic encoding procedure. baseline (sequential): A particular sequential DCT-based encoding and decoding process specified in this binary

29、 decision: Choice between two alternatives. bit stream: Partially encoded or decoded sequence of bits comprising an entropy-coded segment. block: An 8 x 8 array of samples or an 8 x 8 array of DCT coefficient values of one component. 3.1.12 3.1.13 3.1.14 the entropy-coded segment following the gener

30、ation of an encoded hexadecimal XFF byte. block-row: A sequence of eight contiguous component lines which are partitioned into 8 x 8 blocks. byte: A group of 8 bits. byte stumng: A procedure in which either the Huffman coder or the arithmetic coder inserts a zero byte into 3.1.15 overflows the eight

31、 bits reserved for the output byte. 3.1.16 by selecting the smallest integer value which is greater than or equal to the real number. 3.1.17 class (of coding process): Lossy or lossless coding processes. 3.1.18 code register: The arithmetic encoder register containing the least significant bits of t

32、he partially completed cntropy-coded segment. Alternatively, the arithmetic decoder register containing the most significant bits of a partially ilecoded entropy-coded segment. 3.1.19 3.1.2 coding: Encoding or decoding. 3.1.21 3.1.22 3.1.23 3.1.24 3.1.25 3.1.26 3.1.27 3.1.28 3.1.29 interval is great

33、er than the size of the MPS interval (in arithmetic coding). 3.1.30 state machine (in arithmetic coding). 3.1.31 decisions and the conditional probability estimates used in arithmetic coding. 3.1.32 estimation state machine (in arithmetic coding). 3.1.33 3.1.34 processes. carry bit: A bit in the ari

34、thmetic encoder code register which is set if a carry-over in the code register ceiling function: The mathematical procedure in whicb the greatest integer value of a real number is obtained coder: An embodiment of a coding process. coding model: A procedure used to convert input data into symbols to

35、 be coded. (coding) process: A general term for referring to an encoding process, a decoding process, or both. colour image: A continuous-tone image that has more than one component. columns: Samples per line ii7 a component. component: One of the two-dimensional arrays which comprise an image. comp

36、ressed data: Either compressed image data or table specification data or both. compressed image data: A coded representation of an image, as specified in this Specification. compression: Reduction in the number of bits used to represent source image data. conditional exchange: The interchange of MPS

37、 and LPS probability intervals whenever the size of the LPS (conditional) probability estimate: The probability value assigned to the LPS by the probability estimation conditioning table: The set of parameters which select one of the defined relationships between prior coding context: The set of pre

38、viously coded binary decisions which is used to create the index to the probability continuous-tone image: An image whose components have more than one bit per sample. data unit: Ai1 8 ( 8 block of samples of one component in DCT-based processes; a sample in lossless 2 CCII Kec. T.81 (1992 E) ITU-T

39、RECMNxT-BL 93 U 486259L 0586624 O66 ISO/IEC 10918-1 : 1993(E) 3.1.35 3.1.36 previously encoded 8 x 8 block of the same component is subtracted from the current quantized DC coefficient. 3.1.37 coefficient, to a quantized DCT coefficient, or to a dequantized DCT coefficient. 3.1.38 3.1.39 image. DC c

40、oefficient: The DCT coefficient for which the frequency is zero in both dimensions. DC prediction: The procedure used by DCT-based encoders whereby the quantized DC coefficient from the (DCT) coefficient: The amplitude of a specific cosine basis function - may refer to an original DCT decoder: An em

41、bodiment of a decoding process. decoding process: A process which takes as its input compressed image data and outputs a continuous-tone 3.1.40 coding of an image. 3.1.41 DCT coefficients. default conditioning: The values defined for the arithmetic coding conditioning tables at the beginning of dequ

42、antization: The inverse procedure to quantization by which the decoder recovers a representation of the 3.1.42 differential component: The difference between an input component derived from the source image and the corresponding reference component derived from the preceding frame for that component

43、 (in hierarchical mode coding). 3.1.43 decoded. differential frame: A frame in a hierarchical process in which differential components are either encoded or 3.1.44 this Specification. 3.1.45 Specification. 3.1.46 3.1.47 transform. (digital) reconstructed image (data): A continuous-tone image which i

44、s the output of any decoder defined in (digital) source image (data): A continuous-tone image used as input to any encoder defined in this (digital) (still) image: A set of two-dimensional arrays of integer data. discrete cosine transform; DCT: Either the forward discrete cosine transform or the inv

45、erse discrete cosine 3.1.48 mode coding). 3.1.49 3.1.50 data. downsampling (filter): A procedure by which the spatial resolution of an image is reduced (in hierarchical encoder: An embodiment of an encoding process. encoding process: A process which takes as its input a continuous-tone image and out

46、puts compressed image 3.1.51 entropy-coded (data) segment: An independently decodable sequence of entropy encoded bytes of compressed image data. 3.1.52 the entropy encoded segment. 3.1.53 3.1.54 produced by the entropy encoder. 3.1.55 3.1.56 such that the average number of bits per symbol approache

47、s the entropy of the input symbols. 3.1.57 which additional capabilities are added to the baseline sequential process. (entropy-coded segment) pointer: The variable which points to the most recently placed (or fetched) byte in entropy decoder: An embodiment of an entropy decoding procedure. entropy

48、decoding: A lossless procedure which recovers the sequence of symbols from the sequence of hits entropy encoder: An embodiment of an entropy encoding procedure. entropy encoding: A lossless procedure which converts a sequence of input symbols into a sequence of bits extended (DCT-based) process: A d

49、escriptive term for DCT-based encoding and decoding processes in 3.1.58 which converts a block of samples into a corresponding block of original DCT coefficients. forward discrete cosine transform; FDCT: A mathematical transformation using cosine basis functions CCITT Rec. T.81 (1992 E) 3 ITU-T RECMN*T*81 93 4862571 0586625 TT2 = lSO/IEC 10918-1 : 1993(E) 3.1.59 one or more of the components in an image. 3.1.60 are coded at the beginning of a frame. 3.1.61 3.1.62 coding). 3.1.63 mode coding). 3.1.64 grayscale image: A continuous-tone image that has only one component. 3.1