Study and comparison of H.264/MPEG4 part10 AVC main profile with AVS P2 Jizhun profileContentsIntroductionSlide 4H.264 / MPEG 4 AVCCoding structureDifferent YUV systems in H.264 [16]Profiles and levels in H.264H.264 Profiles – coding parts [1]Intra prediction in H.264H.264 intra prediction modes for 4X4 lumaH.264 inter predictionH.264 encoder [1]H.264 decoder [1]De-blocking filterEntropy codingAVS – ChinaAVS Parts [3]AVS layered data structure [3] for 4:2:0 video sequenceAVS video - profilesAVS P2 encoder [ 24]AVS decoderAVS motion estimation and intra prediction:AVS motion estimation and intra prediction (cont..)Predicted pictures (P – pictures):Interpolated Pictures (B-Pictures):AVS intra prediction [7]Comparison H.264 and AVS videoRate control Block:Encoder with and without rate control blockBitrate Vs QPStandards for representing digital videoresults:JM 16.1 outputSlide 35Slide 36AVS ref software outputSlide 38Input sequence : QCIF Foreman [10] : Bitrates Vs psnrPlot: PSNR vs. bit rateSequence: QCIF car phone [10]Slide 42Slide 43CIF: Foreman sequence [10]Slide 45Slide 46CIF football sequence [10]Slide 48Slide 494CIF: sequence ( 704 X 576) [20]Slide 51Slide 52ConclusionFuture work:List of acronyms usedSlide 56ReferencesSlide 58Slide 59Slide 60- By Naveen Siddaraju [email protected] - Under the guidance of Dr K R RaoStudy and comparison of H.264/MPEG4 part10 AVC main profile with AVS P2 Jizhun profileContents Introduction Brief overview of H.264 Brief overview of AVS ChinaComparison Results Conclusion Future workIntroductionH.264 / MPEG 4 AVCLatest video coding standardBasic design architecture is similar to previous MPEG standards Better compression efficiencyCoding structure Picture slices MBs subMBs blocks pixelsI , P and B slices Derived slices SI and SPDifferent YUV systems in H.264 [16]Profiles and levels in H.264 Baseline profile Main profile Extended profile High profilesH.264 Profiles – coding parts [1]Intra prediction in H.264Uses adaptive spatial prediction to reduce spatial redundancy. 4 X 4 luma, 8X8 luma( used in high profile) – 9 modes 16 X 16 luma – 4 modes 4 X 4 chroma - 4 modesH.264 intra prediction modes for 4X4 lumaH.264 inter prediction Uses motion estimation and motion compensation (MC). Used to reduce temporal correlation.H.264 encoder [1]H.264 decoder [1]De-blocking filterIs used to reduce the blocking artifacts.Two sources of the blocking arifacts. - 4 X 4 transform - motion compensated prediction since the filter is present in the loop , it prevents the propagation of the blocking artifacts.Entropy codingquantized values and other syntax elements into actual bits.CAVLC and CABAC for quantized coefficients.Exp- Golomb codes for syntax elements which are not coded by CABAC or CAVLC.AVS – China Latest audio video coding standard from China Standardization includes system, audio, video and digital copyright management. Goal – to achieve coding efficiency with reduced complexity.AVS Parts [3]AVS layered data structure [3] for 4:2:0 video sequenceAVS video - profilesJizhun profile (base profile)Jiben profile (basic profile)Shenzhan profile (extended profile)Jiaqiang profile (enhanced profile)AVS P2 encoder [ 24]AVS decoderAVS motion estimation and intra prediction: AVS uses adaptive modes for motion compensation at the picture layer and the macroblock layer. The modes in picture layer are : Forward prediction from the most recent reference frameForward prediction from the second most recent prediction frameInterpolative prediction between the most recent reference frame and a future reference frame.Intra codingAVS motion estimation and intra prediction (cont..)At the macroblock layer, the modes depend on the picture mode.In Intra pictures, all macroblocks are intra coded.In Predicted pictures, macroblocks may be forward predicted or intra coded. In interpolated pictures, macroblocks may be forward predicted, backward predicted, interpolated or intra coded.Predicted pictures (P – pictures):Interpolated Pictures (B-Pictures):AVS intra prediction [7]Comparison H.264 and AVS videoRate control Block:A rate control algorithm dynamically adjusts encoder parameters to achieve a target bitrate. quantization parameter QPEncoder with and without rate control blockBitrate Vs QPStandards for representing digital videoresults: H.264 – JM 16.1 used [9]AVS – RM09.01[10]JM 16.1 outputJM 16.1 outputJM 16.1 outputAVS ref software outputAVS ref software outputInput sequence : QCIF Foreman [10] : Bitrates Vs psnrPlot: PSNR vs. bit rateSequence: QCIF car phone [10]CIF: Foreman sequence [10]CIF football sequence [10]4CIF: sequence ( 704 X 576) [20]Conclusion H.264 always gives better coding efficiency and compression. Advantages of AVS: - Faster encoder - Reduced complexity - reduced costs. Performance difference is mainly due to absence of complex tools such as CABAC and fewer prediction modes.Future work: comparison between other profiles like extended profiles of AVS and AVC.Comparison between h.264 and other codecs like wmv9 (windows media video), RV10 (real video) etc.List of acronyms usedThank youReferences [1] Soon-kak Kwon et al. “Overview of H.264 / MPEG-4 Part 10 (pp.186-216)”, Special issue on “ Emerging H.264/AVC video coding standard”, J. Visual Communication and Image Representation, vol. 17, pp.183-552, April 2006. [2] A. Puri et al. “Video coding using the H.264/MPEG-4 AVC compression standard”, Signal Processing: Image Communication, vol.19, pp 793-849, Oct 2004.[3] W. GAO et al. “AVS - The Chinese next-generation video coding standard” NAB, Las Vegas, 2004. [4] X. Wang et.al “Performance comparison of AVS and H.264/AVC video coding standards” J. Comput. Sci. & Technol., Vol.21, No.3, pp.310-314 J, May 2006.[5] L. Yu et al. “An Overview of AVS-Video: tools, performance and complexity”, Visual Communications and Image Processing 2005, Proc. of SPIE, vol. 5960, pp.596021, July 31, 2006.[6] D Chang “MPEG 4, H.264 compression standards” ppt, http://mmlab.snu.ac.kr/[7] L. Yu et al. “An area-efficient VLSI architecture for AVS intra frame encoder” Visual Communications and Image Processing 2007, Proc. of SPIE-IS & T Electronic Imaging, SPIE vol. 6508, pp. 650822, Jan. 29,
View Full Document