1、 INTERNATIONAL TELECOMMUNICATION UNION ITU-T J.240TELECOMMUNICATION STANDARDIZATION SECTOR OF ITU (06/2004) SERIES J: CABLE NETWORKS AND TRANSMISSION OF TELEVISION, SOUND PROGRAMME AND OTHER MULTIMEDIA SIGNALS Measurement of the quality of service Framework for remote monitoring of transmitted pictu
2、re signal-to-noise ratio using spread-spectrum and orthogonal transform ITU-T Recommendation J.240 ITU-T Rec. J.240 (06/2004) i ITU-T Recommendation J.240 Framework for remote monitoring of transmitted picture signal-to-noise ratio using spread-spectrum and orthogonal transform Summary This Recommen
3、dation presents an effective framework for remote monitoring of video quality for contribution and primary distribution of digital television transmission. In this framework, image features are extracted at each link point in the transmission chain using spread-spectrum and orthogonal transform. The
4、 extracted image features, i.e., coefficients, are transmitted to the central monitoring room by data circuit that is separated from the video transmission circuit and picture quality is estimated by comparing the coefficients. A framework for extracting the coefficients is presented below. An examp
5、le of coefficient extraction suitable for PSNR estimation is described in Appendix I. The theoretical background for PSNR estimation and the simulation results showing the effectiveness of this framework are also described in Appendix I. Source ITU-T Recommendation J.240 was approved on 29 June 2004
6、 by ITU-T Study Group 9 (2001-2004) under the ITU-T Recommendation A.8 procedure. ii ITU-T Rec. J.240 (06/2004) FOREWORD The International Telecommunication Union (ITU) is the United Nations specialized agency in the field of telecommunications. The ITU Telecommunication Standardization Sector (ITU-
7、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 telecommunications on a worldwide basis. The World Telecommunication Standardization Assembly (WTSA), which meets every four ye
8、ars, 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 laid down in WTSA Resolution 1. In some areas of information technology which fall within ITU-Ts purview, the nec
9、essary 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 telecommunication administration and a recognized operating agency. Compliance with this Recommendation is voluntary. Howev
10、er, 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 mandatory provisions are met. The words “shall“ or some other obligatory language such as “must“ and the negative eq
11、uivalents 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 RIGHTS ITU draws attention to the possibility that the practice or implementation of this Recommendation may involve the use of a
12、 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 others outside of the Recommendation development process. As of the date of approval of this Recommendation, ITU
13、had received notice of intellectual property, protected by patents, which may be required to implement this Recommendation. However, implementors are cautioned that this may not represent the latest information and are therefore strongly urged to consult the TSB patent database. ITU 2004 All rights
14、reserved. No part of this publication may be reproduced, by any means whatsoever, without the prior written permission of ITU. ITU-T Rec. J.240 (06/2004) iii CONTENTS Page 1 Scope 1 2 References. 1 2.1 Normative reference . 1 2.2 Informative references 1 3 Definitions 1 4 Abbreviations 1 5 Picture q
15、uality monitoring based on transform coefficients extraction 2 5.1 Configuration of transmission chain 2 5.2 Extraction of image feature 3 5.3 Clipping, rounding and coding of coefficients. 4 5.4 Additional information . 4 5.5 PSNR estimation 4 Appendix I Implementation example 6 I.1 Feature extract
16、ion . 6 I.2 Coding of coefficients 6 I.3 MSE calculation . 6 I.4 Theoretical backgrounds 7 I.5 Performances 8 ITU-T Rec. J.240 (06/2004) 1 ITU-T Recommendation J.240 Framework for remote monitoring of transmitted picture signal-to-noise ratio using spread-spectrum and orthogonal transform 1 Scope Th
17、is Recommendation presents a framework for automatic remote picture quality monitoring. The proposed framework extracts transform coefficients of the transmitted picture and transmits them to the monitoring operator by an additional data circuit. As the coefficients extraction and transmission is pe
18、rformed separately from the mainstream video transmission, this framework does not affect the quality of transmitted video and thus this is applicable for monitoring video transmission that requires high quality such as contribution and primary distribution. 2 References 2.1 Normative reference The
19、following ITU-T Recommendations and other references contain provisions which, through reference in this text, constitute provisions of this Recommendation. At the time of publication, the editions indicated were valid. All Recommendations and other references are subject to revision; users of this
20、Recommendation are therefore encouraged to investigate the possibility of applying the most recent edition of the Recommendations and other references listed below. A list of the currently valid ITU-T Recommendations is regularly published. The reference to a document within this Recommendation does
21、 not give it, as a stand-alone document, the status of a Recommendation. ITU-T Recommendation J.143 (2000), User requirements for objective perceptual video quality measurements in digital cable television. 2.2 Informative references ITU-T Recommendation J.144 (2004), Objective perceptual video qual
22、ity measurement techniques for digital cable television in the presence of a full reference. ITU-T Recommendation J.147 (2002), Objective picture quality measurement method by use of in-service test signals. ITU-R Recommendation BT.656-4 (1998), Interfaces for digital component video signals in 525-
23、line and 625-line television systems operating at the 4:2:2 level of Recommendation ITU-R BT.601 (Part A). 3 Definitions This Recommendation defines the following term: 3.1 node: Link point in the transmission chain. 4 Abbreviations This Recommendation uses the following abbreviations: DEC Decoder E
24、NC Encoder OT Orthogonal Transform PSNR Peak Signal-to-Noise Ratio 2 ITU-T Rec. J.240 (06/2004) SS Spread Spectrum WHT Walsh-Hadamard Transform 5 Picture quality monitoring based on transform coefficients extraction 5.1 Configuration of transmission chain Configuration of the assumed transmission ch
25、ain is shown in Figure 1. It is assumed that baseband signal is present at each edge (nodes 0 and 1) of the video transmission circuit in which coding and decoding equipment are included. The video signal branches away from the mainstream at nodes 0 and 1 and the branched signals are put into a feat
26、ure extractor. The extracted feature information is transmitted to a PSNR estimator via a data circuit that is separated from the video transmission circuit. The PSNR estimator measures the transmission quality PSNR using feature information at each node. ReceivedvideoVideo transmissionSourcevideoFe
27、atureextractorNode 0Image feature transmission(Data transmission)Video transmissionFeatureextractorPSNRestimatorNode 1MonitoringterminalFigure 1/J.240 Basic configuration Although Figure 1 is configured for monitoring a single link, this can be extended to monitoring of a transmission chain by casca
28、ding the single link as shown in Figure 2. In Figure 2, feature extractors are connected at each node of the transmission chain and the feature information is transmitted to the PSNR estimator. At the PSNR estimator, feature information of two nodes between which transmission quality is monitored is
29、 selected and PSNR is estimated using the selected feature information. ITU-T Rec. J.240 (06/2004) 3 ReceivedvideoENC DEC ENC DEC Source video x nFeatureextractorFeatureextractorFeatureextractorDEC ENC Node 0Node 1FeatureextractorPSNR estimator Node iNode LPicture qualityImage feature transmission(D
30、ata transmission)Video transmissionFigure 2/J.240 Transmission chain monitoring 5.2 Extraction of image feature The image feature is extracted in each pixel block after dividing the input image into blocks of size Nx Ny. When the number of pixels in horizontal and vertical direction is not equal to
31、the multiple of Nxand Ny respectively, additional pixels with a middle pixel value are padded to make the number of pixels equal to Nxand Ny. Figure 3 shows the operation for extracting the image feature. The image feature is obtained by the combination of the spread spectrum and orthogonal transfor
32、m. Let xf(b)(n) and sPN(b)(n) denote the input signal of block b in frame f and PN sequence of block b respectively, where n is the index of pixels. Then, the image feature Rfb is obtained as follows using orthogonal transform OT: () 0Amp kXbRbff= (1) () ()()() ()()()() nsnxOTnxSSOTkXbPNbfbfbf= (2)
33、where SSOT stands for Spread Spectrum and Orthogonal Transform, AmpZ for the amplitude component of Z, and k0for the position of the extracted coefficient. Rfb can also be defined as follows applying inverse orthogonal transform OT 1after multiplying another PN sequence sPN2: () 0Amp nybRbff= (3) ()
34、 ()()() ()()()() kskXOTnxSSOTnybPNbfbfbf212= (4) where n0is the position of the extracted coefficient (in this case, pixel). This second implementation strengthens the effectiveness of spread spectrum by randomizing the spatial domain in addition to the frequency domain. 4 ITU-T Rec. J.240 (06/2004)
35、 InverseorthogonaltransformChoose one coef.andcalculate amplitudeClipping, roundingandcodingGeneratetimecodeMUXOrthogonaltransformxnf()()bRbf Tf t = 0Cb,Tff Current time t0Cbf ()nSPN()b()nSPN2()bFigure 3/J.240 Operation in feature extractor As for PN sequence, different sequences can be used for eac
36、h block. But the same PN sequences must be used at all nodes in the transmission chain for each block in the picture. This is a mandatory requisite for synchronization between nodes, which is described in 5.3. 5.3 Clipping, rounding and coding of coefficients The extracted coefficient Rf b is output
37、ted to the data circuit. As Rfb is generally a real number, clipping and rounding are applied to Rfb in order to express the coefficient by limited bit length. Compression coding may be applied optionally for data compression provided that the probability distribution of Rfb is biased and thus the a
38、mount of information is expected to be reduced. All coded coefficients in a frame or field are packed and data is outputted. 5.4 Additional information As additional information for the PSNR estimator, time information Tf can be outputted at the feature extractor. Time information describes the corr
39、espondence of the frame or field number with the output time of the coefficients, which is multiplexed with coded coefficients Cfb at the final stage of the feature extractor. When a timecode such as LTC and VITC is available, the timecode can be transmitted instead of output time. This is used for
40、delay adjustment of image features in the PSNR Estimator. 5.5 PSNR estimation The operation in the PSNR Estimator is shown in Figure 4. First, multiplexed information sent from two nodes which are located at each edge of the monitored link is demultiplexed and coded coefficients C0fb, C1fb and time
41、information T0f, T1f are obtained. When the order of arrival of the coefficients and the time information does not conform to the order of output at the feature extractor, the received information have to be sorted to ensure that the coefficients and the time information are arranged in frame number
42、 order. Next, in case the coefficients are coded, decoding them is applied and coefficients D0fb and D1fb are obtained. Then delay between two nodes is adjusted using the coefficients and the time information T0f and T1f. T0fT1fF,(D b D b )01f f DelayadjustmentDMUX(From Node 0) Cb,Tf00 f Decoding Cb
43、0f MSEPSNRDMUX(From Node 1) Cb,Tf11 f Decoding Cb1f Db0f Db1f Cb0f Cb1f Figure 4/J.240 Operations in PSNR estimator ITU-T Rec. J.240 (06/2004) 5 The necessity for delay adjustment is shown in Figure 5. There is a video transmission delay dV between two nodes and data transmission delay dC0 and dC1 b
44、etween each node and the PSNR estimator. The coefficients of the frame “F1“ that is displayed on time t1on the transmission side arrives at the PSNR estimator on time t1 + dC0 from Node 0 and t1 + dV+ dC1from Node 1 respectively. Thus, arrival time of the coefficients of two nodes is generally diffe
45、rent and thus it is not guaranteed that the coefficients for the same frame arrive at the PSNR estimator simultaneously from two nodes. Further, some transmission procedures of the data circuit do not ensure that the order of the coefficients reception is the same as the transmission order (i.e., fr
46、ame number order). Therefore, it is meaningless to compare the coefficients that arrive at the PSNR estimator at the same time from two nodes. The PSNR estimator should synchronize the coefficients from two nodes at the “delay adjustment“ block after sorting the coefficients in transmission order at
47、 DMUX output. Time information T0f and T1f can be utilized for this purpose. F0 F1 F2 F3 TransmissionTimet1Node 0 Node 1FeatureextractorF0 F1 F2 F3Timet+d1VCoefficientsof Frame F0Delay in coefficient transmission from Node 1 to PSNR Estimator: dC1Video transmissiondelay : dVt+d1C0t+d+d1VC1Delay in c
48、oefficient transmission from Node 0 to PSNR Estimator: dC0PSNR estimatorFeatureextractorCoefficientsof Frame F1Coefficientsof Frame F2Coefficientsof Frame F0Coefficientsof Frame F1Coefficientsof Frame F2Figure 5/J.240 Delay adjustments between nodes After synchronizing the coefficients, the degradat
49、ion of picture quality caused by video transmission is calculated using the coefficients D0fb and D1fb. Let MSE denote the degradation. It can be expressed as follows: ()210MSE bDbDAverageff= (5) PSNR is derived from MSE as follows: MSE21025520logPSNR = (6) By repeating the above procedure in all frames, the picture quality of the link between Node 0 and Node 1 can be obtained. When PSNR is to be used for subjective quality estimates, it is important
