《ISO-14496-4-CORR-2-2007.pdf》由会员分享,可在线阅读,更多相关《ISO-14496-4-CORR-2-2007.pdf(3页珍藏版)》请在三一文库上搜索。
1、ICS 35.040 Ref. No. ISO/IEC 14496-4:2004/Cor.2:2007(E) ISO/IEC 2007 All rights reserved Published in Switzerland INTERNATIONAL STANDARD ISO/IEC 14496-4:2004 TECHNICAL CORRIGENDUM 2 Published 2007-04-01 INTERNATIONAL ORGANIZATION FOR STANDARDIZATION ORGANISATION INTERNATIONALE DE NORMALISATION INTERN
2、ATIONAL ELECTROTECHNICAL COMMISSION COMMISSION LECTROTECHNIQUE INTERNATIONALE Information technology Coding of audio-visual objects Part 4: Conformance testing TECHNICAL CORRIGENDUM 2 Technologies de linformation Codage des objets audiovisuels Partie 4: Essai de conformit RECTIFICATIF TECHNIQUE 2 Te
3、chnical Corrigendum 2 to ISO/IEC 14496-4:2004 was prepared by Joint Technical Committee ISO/IEC JTC 1, Information technology, Subcommittee SC 29, Coding of audio, picture, multimedia and hypermedia information. In Clause 2, “Normative references”, remove the following: “IEEE Standard Specifications
4、 for the Implementations of 8 by 8 Inverse Discrete Cosine Transform, IEEE Std 1180-1990, December 6, 1990” ISO/IEC 14496-4:2004/Cor.2:2007(E) 2 ISO/IEC 2007 All rights reserved In Clause 2, “Normative references”, add the following: “ISO/IEC 23002-1, Information technology MPEG video technologies P
5、art 1: Accuracy requirements for implementation of integer-output 88 inverse discrete cosine transform” In 5.3, “Procedure for testing bitstream compliance”, replace “A verifier which does perform the IDCT transform and calculates the reconstructed samples must comply with all the arithmetic precisi
6、on requirements specified in ISO/IEC 14496-2. In addition, the IDCT of such a verifier shall be an embodiment of the saturated mathematical integer-number IDCT specified in Annex A of ISO/IEC 14496-2 (a software implementation using 64-bit double-precision floating-point is sufficient).” with the fo
7、llowing: “A verifier which does perform the IDCT transform and calculates the reconstructed samples must comply with all the arithmetic precision requirements specified in ISO/IEC 14496-2. In addition, the IDCT of such a verifier shall be an embodiment of the mathematical integer-number IDCT specifi
8、ed in Annex A of ISO/IEC 14496-2:2004 (a software implementation using 64-bit double-precision floating-point is sufficient).” In 5.4, “Definition of visual decoder compliance”, replace “The reference decoder is a decoder that implements precisely the decoding process as specified in ISO/IEC 14496-2
9、. The IDCT function that shall be used when running the reference decoder is the very accurate approximation of the mathematical saturated integer-number IDCT specified in Annex A of ISO/IEC 14496-2 obtained by implementing IDCT with double-precision arithmetic.” with the following: “The reference d
10、ecoder is a decoder that implements precisely the decoding process as specified in ISO/IEC 14496-2. The IDCT function that is used when running the reference decoder shall be an embodiment of the mathematical integer-number IDCT specified in Annex A of ISO/IEC 14496-2:2004 (a software implementation
11、 using 64-bit double-precision floating-point is sufficient).” Replace the content of 5.4.2, “Requirement on arithmetic accuracy in video objects (with IDCT)”, with the following: “When a bitstream contains some 8x8 blocks with non-zero DCT coefficients, the output of a compliant decoder may differ
12、from the output of the reference decoder. However, because of the accuracy requirements on the IDCT implementation used by the decoder, there exist some accuracy requirements on the output of a compliant ISO/IEC 14496-2 video decoder. The IDCT implementation used in a compliant decoder shall meet al
13、l the requirements defined in Annex A of ISO/IEC 14496-2:2004, including all tests specified therein.” In 5.5.3, “Specification of the test bitstreams”, replace “All the bitstreams in the test suite must be such that the output of the non-saturated integer number mathematical IDCT, as defined in Ann
14、ex A of ISO/IEC 14496-2, has values within the range -384, 383 for each coded block.” with the following: “All the bitstreams in the test suite must be such that the output of the mathematical integer-number IDCT, as defined in Annex A of ISO/IEC 14496-2:2004, has values within the range -384, 383 f
15、or each coded block.” -,-,- ISO/IEC 14496-4:2004/Cor.2:2007(E) ISO/IEC 2007 All rights reserved 3 Replace the content of 5.5.3.1.15, “Test bitstream #GE-19”, with the following: “Specification: A bitstream implementing a test close to the ISO/IEC 23002-1 main body IDCT mismatch test, to test the dec
16、oders IDCT statistical accuracy. The test can be performed using I-VOPs with quant_type equal to 1, a flat custom quantisation matrix with all entries Wwvu equal to 8, and quantiser_scale equal to 1. Use whatever number of VOPs are required to satisfy a statistic count corresponding to the requireme
17、nts of Annex A of ISO/IEC 14496-2:2004 for use in ISO/IEC 23002-1 IDCT testing. Note that because of the saturation range of 0, 255 in the video decoding process, the test cannot exactly emulate the ISO/IEC 23002-1 main body IDCT test. Functional stage: IDCT Purpose: Check IDCT decoder accuracy. Thi
18、s is not a drift test since all macroblocks are of type Intra.” Replace the content of 5.5.3.1.18, “Test bitstream #GE-22”, with the following: “Specification: Bitstream causing large positive sample domain coefficients fyx (e.g., 255) added to large predicted values pyx (e.g., 255), or large negati
19、ve sample domain coefficients fyx (e.g., -256) added to small predicted values pyx (e.g., 0). Functional stage: addition of the output of IDCT fyx to the predicted values pyx and saturation of the result to the range 0, 2n 1. Purpose: Test that decoder implements properly the addition of the output
20、of IDCT fyx to the predicted values pyx and saturation of the result to the range 0, 2n 1.” Replace the content of 5.5.3.1.21, “Test bitstream #GE-25”, with the following: “Specification: A bitstream in which the output of the mathematical integer-number IDCT f (x, y), as defined in Annex A of ISO/I
21、EC 14496-2, has large absolute values within the range -2n-2n-1, 2n+2n-1-1 for each coded block, where n is the maximum allowed number of bits per pixel for the profile-and-level combination. Functional stage: IDCT Purpose: Check that IDCT decoder accuracy meets the requirements defined in Annex A o
22、f ISO/IEC 14496-2:2004. The peak absolute error for a compliant decoder shall be less than or equal to 2 when decoding this bitstream. Note that for blocks where f (x, y) has values within the range -300, 300, decoders that have a peak absolute error larger than 1 may not be compliant with the main body of ISO/IEC 23002-1.” -,-,-