ETSI ETS 300 142-1998 Integrated Services Digital Network (ISDN) and Other Digital Telecommunications Networks Line Transmission of Non-Telephone Signals Video Codec for Audiovisua_1.pdf

1、STANDARD ETS 300 142 June 1998 Source: NA Reference: DEMA-051 108 ICs: 33.020 Key words: Multichannel, codec, telesetvice, ISDN, video Integrated Services Digital Network (ISDN) and other digital telecommunications networks; Line transmission of non-telephone signals; Video codec for audiovisual ser

2、vices at p x 64 kbit/s ITU-T Recommendation H.261 (1993), modified ETSI Eu ropean Telecom mun cat ions Standards Institute ETSI Secretariat Postal address: F-06921 Sophia Antipolis CEDEX - FRANCE Office address: 650 Route des Lucioles - Sophia Antipolis - Valbonne - FRANCE Internet: secretariatQetsi

3、.fr - http:/www.etsi.fr - http:/www.etsi.org Tel.: 43 4 92 94 42 O0 - Fax: 43 4 93 65 47 16 Copyright Notification: No pari may be reproduced except as authorized by written permission. The copyright and the foregoing restriction extend to reproduction in all media. O European Telecommunications Sta

4、ndards Institute 1998. All rights reserved. Page 2 ETS 300 142: June 1998 Whilst every care has been taken in the preparation and publication of this document, errors in content, typographical or otherwise, may occur. If you heve oomments concerning its accuracy, please write to ETSI Standards Makin

5、g Support Dept at the address shown on the Utle page. STDoETSI ETS 300 142-ENGL 1938 m 3400855 0311432 585 m Page 3 ETS 300 142: June 1998 Foreword This European Telecommunication Standard (ETS) has been produced by the Network Aspects (NA) Technical Committee of the European Telecommunications Stan

6、dards Institute (ETSI). Transposition dates Date of adoption of this ETS: Date of latest announcement of this ETS (doa): 19 June 1998 30 September 1998 Date of latest publication of new National Standard or endorsement of this ETS (doNe): 31 March 1999 Date of withdrawal of any conflicting National

7、Standard (dow): 31 March 1999 Endorsement notice The text of ITU-T Recommendation H.261 (1993) was approved by ETSI as an ETS with agreed modifications as given below. NOTE: New or modified text is indicated using sidebars. In addition, underlining and/or strike- out are used to highlight detailed m

8、odifications where necessary. Page 4 ETS 300 142: June 1998 Page 1 Replace the scope clause with the following two clauses (scope and normative references): 1 Scone mETS describes the video cod na and decod na methods for the movina Dicture comDonent o f audiovisual services at the rates of D x 64 I

9、d3 iVs. where D is in the ranae 1 to 30. The standard includes CO nformance test sDecifications. Normative references This ETS incorporates bv dated and undated reference. Drovisions from other oubiications. These re lised normative references are cited at the am rom iate daces in the text and the D

10、ub lications a hereafter. For dated references. subseaue nt ame ndments to or rev isions of anv of these etions amlv to this ETS onlv when incomorated in it bv amendments or revision. For undated references the latest edition of the oublication referred to amlies. ITU-T Recommendation H 120 Codecs f

11、o r videoconfe rencina us na Drimaw diaital arour, transmission. ITU-T Recomme ndation H.200: Framewo rk for Recomrnenda tions fo r audiovisual services. ITU-T Recommendation H.221: Frame structure for a 64 to 1 920 kb ius channel in audiovisual telesew ices. ITU-T RSO m m endation H .230: Fram e-sv

12、nc hro nous cont rol and indicati on gianals for audiovisual svstems. ITU-T Recommendation H 742: Svstem for establ ishina commun ication between aud iovisual terminals us ing diaital cha nnels UD to 2 Mbis . ITU-T Recommendation T.81: Information technoloov - Riaital cornmession and codina of CO nt

13、inuous-tone still imaaes - Reau irements and .I I1 I 121 131 141 151 *I idelines. 161 Page 19, subclause 59, second and third paragraphs When operating with CIF the number of bits created by coding any single picture shall not exceed When operating with QCIF the number of bits created by coding any

14、single picture shall not exceed Page 25, clause D.5, third bullet I 256 IK = 1 024) . I 64kbibJ-* - the maximum number of bits allwed per frame (sub-image) should not be exceeded I (256 Dius for CIF and 64 kbis for QCIF, where K is eaual to 1 024 1; Addition of annex ZA Add a new annex 24 as follows

15、: Page 5 ETS 300 142: June 1998 Annex ZA (normative): Conformance testing of ITU-T Recommendation H.261 (March 1993) ZA.l General principles adopted for the ITU-T Recommendation H.261 conformance tests The conformance test should aim to test for compliance to the respective ETS to the extent which i

16、t is necessary to ensure compatible interworking of different suppliers equipment. The only inputs and outputs of the equipment under test in the conformance verification are those normaiiy provided to the user and the test shall not rely on directly accessing additional internal test modes in the e

17、quipment. The specification is constrained to only testing the ITU-T Recommendation H.261 bitstream which can be assumed to have been correctly isolated from the ITU-T Recommendation H.221 framing, audio, etc. The tests should be constructed to allow for devices which only have visual inputs and out

18、puts (e.g. a video phone). The existence of composite (or other) electrical interfaces cannot be assumed. If electrical video inputs and outputs do exist (e.g. composite interfaces) then they should be fully tested in respect of bansmittedreceived active picture area, etc. The testing of decoders is

19、 likely to be more exhaustive than testing of the encoders due to the fact that decoders will always have a digital channel input Assuming that the encoder and decoder of the same manufacturer perform well together, then advantages will be taken of sophisticated decoder tests. An indication of the c

20、orrect behaviour of the decoder, will likely implicate that the complementary part in the encoder is correct as well. Examples of input sequences (video stimulus etc.) will be given as annexes to the conformance specification. These examples will be given to aid the construction of conformance test

21、suites. Tests will not be constrained to only using these examples. Table ZA.l: List of tests Items Encoder Decoder 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Video levels Geometrical distortion Frame rates QCIF support Valid motion vectors Loop filter Inverse DCT Forced updating Video multi

22、plex Video demultiplex Scanning order Freeze picture Fast update request Maximum Bits per picture Error correction FEC stuffing (optional) Arithmetic loop conformance HRD compliance Extra Insertion Information Clipping E E E E E E E E E E E E E D D D D D D D D D D D D D D D D , .- - STDoETSI ETS 300

23、 142-ENGL 1998 3400855 0331435 294 Page 6 ETS 300 142: June 1998 ZA.2 Description of tests ZA.2.1 PURPOSE: Video levels (ITU-T Recommendation H.261 Q 3.1) As the way to produce digital input signals is not a part of ITU-T Recommendation H.261, the only purpose of this test is to verify that the enco

24、der supports input values from 1 to 254 for luminance and the two colour difference components. This test can only be performed when digital ID are available. A complete test of the video levels should take place in a complete system quality conformance (black leve Vw h i te I evelA in ea rity). ZA.

25、21.1 Encoder METHOD: Connect a digital source pattern which contains all the permitted values to the encoder under test and connect the output of the encoder under test to the reference decoder. RESULTS: Check the pictures decoded with the reference decoder. Inputs from black to white shall imply ou

26、tputs to black to white. ZA.21.2 Decoder METHOD: Encode the same digital source pattern with the reference encoder and decode the produced bitstream with the decoder under test. Digital source pattern: uniform picture with programmable values Y, CB and CR from 1 to 254. RESULTS: Check the pictures d

27、ecoded with the decoder under test. Pattern values corresponding to black to white shall imply outputs to black to white. Value 128 shall give grey picture. ZA.2.2 Geometrical Distortion (ITU-T Recommendation H.261 5 3.1) ZA.221 Encoder PURPOSE: A CIF encoder shall take luminance and colour differen

28、ce samples in accordance with the requirements of ITU-T Recommendation H261 5 3.2 and shall maintain an aspect ratio of 4:3. The puipose of this test is to verify that sampling and aspect ratio. METHOD: If the encoder under test is capable of operating at CIF, then, with the encoder operating at tha

29、t nrsolution, provide an image containing a square, circle and crms all centred on the centre of the image (see figure ZA.1). Connect the output of the encoder under test to the reference decoder. RESULT: Check the pictures decoded with the reference decoder. Verify that there has been no geometrica

30、l distortion of the image. STD-ETSI ETS 300 142-ENGL 1998 D 3400855 031143b 120 Page 7 ETS 300 142: June 1998 Figure A.l ZA.2.3 Frame rates (ITU-T Recommendation H.261 9 3.1) ZA.23.1 Encoder PURPOSE: An encoder shall control its output bitstream to comply with the requirements of the Hypothetical Re

31、ference Decoder (HRD) defined in ITU-T Recommendation H.261, annex B. In ITU-T Recommendation H.261 !j 3.1 is stated: “Means shallbeprovidedto restrict the maximum picture rate of encoders by having at least O, 1, 2 or 3 non-transmitted pictures between transmitted ones. Selection of this minimum nu

32、mber and CIF or QCIF shall be by external means (for example via ITU- T Recommendation H.221) : The purpose of the test is to verify that an encoder never sends more pictures per second than signalled by the decoder. METHOD: At the encoder provide an image sequence which contains the following, low

33、detail stills, long periods of high motion, panning, zooming, scene cuts, and high detail stills. The output bitstream (ITU-T Recommendation H.261) is stored for off-line analysis. A minimum of 55 minutes of picture date should be stored. NOTE: To reduce storage requirements the following could be s

34、tored instead: - - temporal reference count; - accurate time stamps for the Picture Start Codes (PSCs); the number of bits between each PSC. All allowable bit rates which are within the encoders declared range should be tested. Input this video sequence to an encoder. Signal to the encoder the decod

35、er capabilities of O, 1, 2, 3 non transmitted pictures between transmitted ones. Supervise the Temporal Reference in the picture layer. Supervise the hypothetical buffer verifier. Perform the test for QCIF and CIF (if the encoder claims to encode also CIF). Page 8 ETS 300 142: June 1998 RESULTS: An

36、encoder can only be said to be compliant if the mean rate of Picture Start Codes is 29,97 Hz f 50 ppm and it satisfies the HRD. a) The exchange of decoder capabilities (ITU-T Recommendations H.221 and H.242) results in allowed values for the Temporal Reference (TR) in the picture layer. The allowed

37、maximum encoder picture rate is (29,97 Hz f 50 ppm) / (n+l); n = O, 1,2,3. Then the corresponding modulo 32 difference between an actual TR and the TR of the last encoded picture can have all integer values between n+l, n+2, . 31. In all cases the buffer of the Hypothetical Reference Decoder (ITU-T

38、Recommendation H.261, annex 6) shall not overflow (the data stream shall conform to the hypothetical buffer verifier). Signalled is n = 2; then the maximum allowed picture rate is 29,97 Hz divided by 3 (approximately 10 Hz). The modulo 32 difference between an actual TR and the TR of the last receiv

39、ed picture can have all values between 3 and 31. b) EXAMPLE: ZA.23.2 Decoder PURPOSE: Purpose of the test is to verify that a decoder which claims to have capability to decode sequences with encoded picture rates of: (29,97 Hz i: 50 ppm) / (n+l); n = O, 1,2,3 really does it for all picture material.

40、 METHOD: Input a data stream conforming to ITU-T Recommendation H.261 describing a video sequence (as the one used in subclause ZA.2.3.1) to a decoder at the lowest claimed value of n for at least 55 minutes. Care should be taken to ensure the data stream describes a sequence encoded at the claimed

41、value of n (Le. there is no picture drcpping). RESULTS: The decoded pictures should be the same as on the reference decoder. EXAMPLE: Most critical should be, e.g. the following sequence: - a synthetic picture sequence consisting of pictures with diagonal stripes (high saturation, diagonal frequency

42、 comrsponding to approx. 1 MHz), the stripes being unmoved for some time and then being moved with 1 pel per encoded frame both horizontally and vettically. STDmETSI ETS 300 LY2-ENGL 1996 3400855 03LL43B TT3 Page 9 ETS 300 142: June 1998 ZA.2.4 QCIF Support (ITU-T Recommendation H.261 5 3.1) PURPOSE

43、: The ITU-T Recommendation H.261 indicates that two formats are allowed as input to the video coder: the Common Intermediate Format (CIF) and another format known as Quarter-CIF (QCIF) having half the number of pel and half the number of lines of the former. In the first format (CIF), the luminance

44、sampling structure is 352 pels per line, 288 lines per picture in an orthogonal arrangement. Sampling of each of the two colour difference components is at 144 lines, 176 pixels per line, orthogonal. ITU-T Recommendation H.261 also specifies that the picture area covered by this number of pixels and

45、 lines has an aspect ratio of 4:3. In addition ITU-T Recommendation H.261 also states that all codecs shall be able to operate using QCIF. This statement prevents the possibility for an encoder to always transmit the entire picture in the CIF form at and leaving to a decoder, only able to operate on

46、 QCIF, the task of selecting from the video stream the portion corresponding to its need. From this consideration the need arises of testing the QCIF support for both encoder and decoder. In fact when an encoder operates with the QCIF format the decoder shall be able to show a correct picture and co

47、nversely if a decoder is only supporting the QCIF the encoder shall be able to transmit a coded bitstream containing only the information related to a QCIF picture. ZA.24.1 Encoder METHOD: The encoder is set in QCIF mode and the video encoder output is connected to a testing system implementing a re

48、ference decoder (either real time or not) enriched by the capability of analysing the bitstream at the Group Of Block (GOB) layer and at the picture layer. Bit 4 of the type information in the picture layer (PTYPE) may only have the value of O. The range of permitted values for the Group Number (GN)

49、 is limited to 1, 3, 5. Other values of PTYPE or GN are not allowed. Other features of the coded and displayed images shall be checked through the conformance testing involving the active picture area and the block structure. RESULTS: If analyzing the values of the GN contained in the coded bitstream, the presence of values different from 1,3 or 5 is detected, this certainly reveals a non correct coding of the QCIF picture. If the image obtained by a reference decoder which is decoding any image coming from the coder under test, doesnt appear as a rect