ITU-T J 343 AMD 1-2018 Hybrid perceptual bitstream models for objective video quality measurements Amendment 1 (Study Group 12).pdf

1、 I n t e r n a t i o n a l T e l e c o m m u n i c a t i o n U n i o n ITU-T J.343 TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU Amendment 1 (05/2018) SERIES J: CABLE NETWORKS AND TRANSMISSION OF TELEVISION, SOUND PROGRAMME AND OTHER MULTIMEDIA SIGNALS Measurement of the quality of service - Part

2、3 Hybrid perceptual bitstream models for objective video quality measurements Amendment 1 Recommendation ITU-T J.343 (2014) Amendment 1 Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) i Recommendation ITU-T J.343 Hybrid perceptual bitstream models for objective video quality measurements Amendment 1 Summary

3、 Recommendation ITU-T J.343 specifies objective video quality measurement methods which use bitstream data in addition to processed video sequences. From bitstream data, the models can obtain additional information on the codec type, bit rate, frame rate, some transmission errors and spatial/tempora

4、l shifts. Consequently, such models may provide improved performance compared to objective video quality models, which use only processed video sequences. Amendment 1 adds a new Appendix I containing information about test vectors for the ITU-T J.343 family of Recommendations. History Edition Recomm

5、endation Approval Study Group Unique ID* 1.0 ITU-T J.343 2014-11-29 9 11.1002/1000/12315 1.1 ITU-T J.343 (2014) Amd. 1 2018-05-10 12 11.1002/1000/13619 Keywords Bitstream data, objective methods, test vectors, video. _ * To access the Recommendation, type the URL http:/handle.itu.int/ in the address

6、 field of your web browser, followed by the Recommendations unique ID. For example, http:/handle.itu.int/11.1002/1000/11830-en. ii Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) FOREWORD The International Telecommunication Union (ITU) is the United Nations specialized agency in the field of telecommunicati

7、ons, information and communication technologies (ICTs). The ITU Telecommunication Standardization Sector (ITU-T) is a permanent organ of ITU. ITU-T is responsible for studying technical, operating and tariff questions and issuing Recommendations on them with a view to standardizing telecommunication

8、s on a worldwide basis. The World Telecommunication Standardization Assembly (WTSA), which meets every four years, establishes the topics for study by the ITU-T study groups which, in turn, produce Recommendations on these topics. The approval of ITU-T Recommendations is covered by the procedure lai

9、d down in WTSA Resolution 1. In some areas of information technology which fall within ITU-Ts purview, the necessary standards are prepared on a collaborative basis with ISO and IEC. NOTE In this Recommendation, the expression “Administration“ is used for conciseness to indicate both a telecommunica

10、tion administration and a recognized operating agency. Compliance with this Recommendation is voluntary. However, the Recommendation may contain certain mandatory provisions (to ensure, e.g., interoperability or applicability) and compliance with the Recommendation is achieved when all of these mand

11、atory provisions are met. The words “shall“ or some other obligatory language such as “must“ and the negative equivalents are used to express requirements. The use of such words does not suggest that compliance with the Recommendation is required of any party. INTELLECTUAL PROPERTY RIGHTSITU draws a

12、ttention to the possibility that the practice or implementation of this Recommendation may involve the use of a claimed Intellectual Property Right. ITU takes no position concerning the evidence, validity or applicability of claimed Intellectual Property Rights, whether asserted by ITU members or ot

13、hers outside of the Recommendation development process. As of the date of approval of this Recommendation, ITU had received notice of intellectual property, protected by patents, which may be required to implement this Recommendation. However, implementers are cautioned that this may not represent t

14、he latest information and are therefore strongly urged to consult the TSB patent database at http:/www.itu.int/ITU-T/ipr/. ITU 2018 All rights reserved. No part of this publication may be reproduced, by any means whatsoever, without the prior written permission of ITU. Rec. ITU-T J.343 (2014)/Amd.1

15、(05/2018) iii Table of Contents Page 1 Scope . 1 1.1 Applications 1 2 References . 1 3 Definitions 2 3.1 Terms defined elsewhere 2 3.2 Terms defined in this Recommendation . 2 4 Abbreviations and acronyms 3 5 Conventions 3 6 Description of hybrid perceptual bitstream model types 4 Annex A Summary of

16、 VQEG validation of hybrid models 7 A.1 Subjective datasets 7 A.2 Model performance summary . 9 Appendix I Test vectors for the ITU-T J.343 family of standards 18 Bibliography. 19 iv Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) Introduction Generally video quality estimation models are, depending upon th

17、e required input signals, categorized as no reference (NR), reduced reference (RR) and full reference (FR) models. NR models are provided with the processed video sequences only. RR models require that features extracted from the reference video sequences and the processed video sequences are provid

18、ed. For FR models, the unimpaired reference and the processed video sequences must be provided. In addition, the models described in this Recommendation need access to the received bitstream data from which the model can obtain information on transmission errors (e.g., delay, packet loss), codec (e.

19、g., type, bit-rates, frame rates, codec parameters), etc. Consequently, the models described here are categorized as hybrid models (i.e., Hybrid-NR, Hybrid-RR and Hybrid-FR). Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) 1 Recommendation ITU-T J.343 Hybrid perceptual bitstream models for objective video q

20、uality measurements Amendment 1 Editorial note: This is a complete-text publication. Modifications introduced by this amendment are shown in revision marks relative to Recommendation ITU-T J.343 (2014). 1 Scope This Recommendation describes recommended objective models for non-intrusive monitoring o

21、f the video quality of IP-based video services based on the decoded video frames and packet-header. Some types of models also utilize the reference video or bitstream information. This Recommendation addresses six application areas: ITU-T J.343.1 specifies Hybrid-NRe models ITU-T J.343.2 specifies H

22、ybrid-NR models ITU-T J.343.3 specifies Hybrid-RRe models ITU-T J.343.4 specifies Hybrid-RR models ITU-T J.343.5 specifies Hybrid-FRe models ITU-T J.343.6 specifies Hybrid-FR models ITU-T J.343.1 includes two models, ITU-T J.343.2 includes one model, ITU-T J.343.3 includes one model that operates at

23、 multiple side channel bandwidths to transmit the reduced reference information, ITU-T J.343.4 includes one model that operates at multiple side channel bandwidths to transmit the reduced reference information, ITU-T J.343.5 includes two models and ITU-T J.343.6 includes two models. All of these mod

24、els predict video quality in terms of mean opinion score (MOS), for example on a five-level absolute category rating (ACR) scale (see ITU-T P.910 or ITU-T P.913). 1.1 Applications This Recommendation describes models that estimate perceptual video quality. The applications for the estimation models

25、described in this Recommendation include, but are not limited to: real-time, in-service quality monitoring at the source; remote destination quality monitoring; quality measurement of transmission systems that utilize video compression and decompression techniques, including concatenations of such t

26、echniques. More information about applications can be found in the individual Recommendations that address these six application areas. 2 References The following ITU-T Recommendations and other references contain provisions which, through reference in this text, constitute provisions of this Recomm

27、endation. At the time of publication, the editions indicated were valid. All Recommendations and other references are subject to revision; users of this Recommendation are therefore encouraged to investigate the possibility of applying the most recent edition of the Recommendations and other referen

28、ces listed below. A list of the currently 2 Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) valid ITU-T Recommendations is regularly published. The reference to a document within this Recommendation does not give it, as a stand-alone document, the status of a Recommendation. ITU-T J.340 Recommendation ITU-T

29、 J.340 (2010), Reference algorithm for computing peak signal to noise ratio of a processed video sequence with compensation for constant spatial shifts, constant temporal shift, and constant luminance gain and offset. ITU-T J.343.1 Recommendation ITU-T J.343.1 (2014), Hybrid-NRe objective perceptual

30、 video quality measurement for HDTV and multimedia IP-based video services in the presence of encrypted bitstream data. ITU-T J.343.2 Recommendation ITU-T J.343.2 (2014), Hybrid-NR objective perceptual video quality measurement for HDTV and multimedia IP-based video services in the presence of non-e

31、ncrypted bitstream data. ITU-T J.343.3 Recommendation ITU-T J.343.3 (2014), Hybrid-RRe objective perceptual video quality measurement for HDTV and multimedia IP-based video services in the presence of a reduced reference signal and encrypted bitstream data. ITU-T J.343.4 Recommendation ITU-T J.343.4

32、 (2014), Hybrid-RR objective perceptual video quality measurement for HDTV and multimedia IP-based video services in the presence of a reduced reference signal and non-encrypted bitstream data. ITU-T J.343.5 Recommendation ITU-T J.343.5 (2014), Hybrid-FRe objective perceptual video quality measureme

33、nt for HDTV and multimedia IP-based video services in the presence of a full reference signal and encrypted bitstream data. ITU-T J.343.6 Recommendation ITU-T J.343.6 (2014), Hybrid-FR objective perceptual video quality measurement for HDTV and multimedia IP-based video services in the presence of a

34、 full reference signal and non-encrypted bitstream data. ITU-T P.910 Recommendation ITU-T P.910 (2008), Subjective video quality assessment methods for multimedia applications. ITU-T P.913 Recommendation ITU-T P.913 (2014), Methods for the subjective assessment of video quality, audio quality and au

35、diovisual quality of Internet video and distribution quality television in any environment. 3 Definitions 3.1 Terms defined elsewhere This Recommendation uses the following terms defines elsewhere: 3.1.1 processed ITU-T P.913: The reference stimuli presented through a system under test. 3.1.2 proces

36、sed video sequence ITU-T P.913: The processed video sequence (PVS) is the impaired version of a video sequence. 3.1.3 reference ITU-T P.913: The original version of each source stimulus. This is the highest quality version available of the audio sample, video clip or audiovisual sequence. 3.2 Terms

37、defined in this Recommendation This Recommendation defines the following terms: 3.2.1 hybrid full reference model: An objective video quality model that predicts subjective quality using the reference video, the decoded video frames, packet headers, and the video payload. Such models cannot analyse

38、encrypted video. Rec. ITU-T J.343 (2014)/Amd.1 (05/2018) 3 3.2.2 hybrid full reference encrypted model: An objective video quality model that predicts subjective quality using the reference video, the decoded video frames, and packet headers. Such models are suitable for use with encrypted video. 3.

39、2.3 hybrid no reference model: An objective video quality model that predicts subjective quality using the decoded video frames, packet headers, and video payload. Such models can be deployed in-service but cannot analyse encrypted video. 3.2.4 hybrid no reference encrypted model: An objective video

40、 quality model that predicts subjective quality using the decoded video frames and packet headers. Such models can be deployed in-service and are suitable for use with encrypted video. 3.2.5 hybrid reduced reference model: An objective video quality model that predicts subjective quality using the d

41、ecoded video frames, packet headers, video payload and features extracted from the reference video. Such models can be deployed in-service but cannot analyse encrypted video. 3.2.6 hybrid reduced reference encrypted model: An objective video quality model that predicts subjective quality using the d

42、ecoded video frames, packet headers, and features extracted from the reference video. These models can be deployed in-service and are suitable for use with encrypted video. 4 Abbreviations and acronyms This Recommendation uses the following abbreviations and acronyms: ACR Absolute Category Rating CO

43、DEC Coder-Decoder ES Elementary bitstream FR Full Reference Hybrid-FR Hybrid Full Reference Hybrid-FRe Hybrid Full Reference encrypted Hybrid-NR Hybrid No Reference Hybrid-NRe Hybrid No Reference encrypted Hybrid-RR Hybrid Reduced Reference Hybrid-RRe Hybrid Reduced Reference encrypted MOS Mean Opin

44、ion Score MPEG Moving Picture Experts Group NR No (or zero) Reference PES Packetized Elementary bitstream PSNR Peak Signal to Noise Ratio PVS Processed Video Sequence RMSE Root-Mean Square Error RR Reduced Reference SRC Source Reference Channel or Circuit VQEG Video Quality Experts Group 4 Rec. ITU-

45、T J.343 (2014)/Amd.1 (05/2018) 5 Conventions None. 6 Description of hybrid perceptual bitstream model types This Recommendation specifies objective video quality measurement methods which use both processed video sequences and bitstream data. The bitstream data may be provided in the forms of elemen

46、tary bitstream (ES), packetized elementary bitstream (PES) or packet video (Figure 1). Table 1 shows required inputs for each model. Table 1 Required inputs Model type Model name Required inputs Hybrid NRe RST-V model YHyNRe Processed video sequence (PVS) Encrypted bitstream Hybrid NR YHyNR PVS Non-

47、encrypted bitstream Hybrid RRe YHyRRe PVS Features extracted from source reference channel (SRC) Encrypted bitstream Hybrid RR YHyRR PVS Features extracted from SRC Non-encrypted bitstream Hybrid FRe PEVQ-S (e) YHyFRe PVS SRC Encrypted bitstream Hybrid FR PEVQ-S YHyFR PVS SRC Non-encrypted bitstream

48、 Hybrid-NR and Hybrid-NRe models use only PVS and bitstream data, as shown in Figure 1 and Figure 2. Where Hybrid-NR models have access to all of this data, Hybrid-NRe models do not have access to the video payload. Therefore, these models can be used with encrypted bitstreams. Rec. ITU-T J.343 (201

49、4)/Amd.1 (05/2018) 5 Figure 1 Block-diagram depicts the core concept of hybrid perceptual bitstream models MOSp: predicted MOS by the model Figure 2 Block-diagram of the Hybrid-NR model In addition to the data available to a Hybrid-NR model, Hybrid-RR and Hybrid-RRe models also use features extracted from source video sequences. Figure 3 shows a Hybrid-RR model. In addition to the bitstream data, the Hybrid-RR model uses the features extracted from the SRC. Where H