AVS-M编码器算法研究和解码器DSP移植和优化
摘要
AVS(Audio Video Coding Standard)标准是我国自主制定的数字音视频编解码技术国家标准。其中AVS的第七部分——移动视频(AVS-M)标准是为了数字存储媒体、网络流媒体、多媒体通信等应用中,针对运动图像压缩技术的需要而制定的。本文首先分析了课题背景和发展现状,简单比较了几种宏块分割模式算法的特点和解码器DSP优化方法。
然后分析研究了AVS-M编解码器的结构和核心算法,将改进的人眼对比灵敏度模型应用到编码器的帧间预测宏块模式选择算法中,得到该算法的快速算法并在PC平台上验证,使编码器在保证视觉质量的前提下,速度提高了53%。
本文又将解码器源码wm2.4移植到TMS320C6416 DSP平台之上,在正常运行的前提下,应用C语言优化、汇编优化、存储器优化和本设计提出的针对插值算法结构的优化这几种方法,使QCIF格式的最终解码速度是优化前的3倍多,达到46帧/秒,满足了实时要求。
AVS (Audio Video Coding Standard)is the standard of video and audio designed by our country. Seven parts of AVS–Mobile Video (AVS-M) is designed for compressing moving images of digital storage, network stream, multimedia communication.
In this paper the background and the status of the research are analyzed firstly. Several algorithms of the choice of macro block partition modes and methods of DSP optimization are showed briefly.
Then structure and algorithms of the AVS-M encoder and decoder are analyzed .In the encoder, the vision model is used in the algorithms of the choice of macro block partition modes. The fast algorithm is applied in encoder on PC ,the speed has been improved by 53% and the effect is good.
Then the decoder source code-wm2.4 is repotted on TMS320C6416 DSP . The optimizations of C code, assembly language , storage space and the structure of interpolation algorithm are used to the decoder, and the results of every method of optimization have been got. As to QCIF the final result is three times as fast as the repotted one. It has met the real time need.
然后分析研究了AVS-M编解码器的结构和核心算法,将改进的人眼对比灵敏度模型应用到编码器的帧间预测宏块模式选择算法中,得到该算法的快速算法并在PC平台上验证,使编码器在保证视觉质量的前提下,速度提高了53%。
本文又将解码器源码wm2.4移植到TMS320C6416 DSP平台之上,在正常运行的前提下,应用C语言优化、汇编优化、存储器优化和本设计提出的针对插值算法结构的优化这几种方法,使QCIF格式的最终解码速度是优化前的3倍多,达到46帧/秒,满足了实时要求。
AVS (Audio Video Coding Standard)is the standard of video and audio designed by our country. Seven parts of AVS–Mobile Video (AVS-M) is designed for compressing moving images of digital storage, network stream, multimedia communication.
In this paper the background and the status of the research are analyzed firstly. Several algorithms of the choice of macro block partition modes and methods of DSP optimization are showed briefly.
Then structure and algorithms of the AVS-M encoder and decoder are analyzed .In the encoder, the vision model is used in the algorithms of the choice of macro block partition modes. The fast algorithm is applied in encoder on PC ,the speed has been improved by 53% and the effect is good.
Then the decoder source code-wm2.4 is repotted on TMS320C6416 DSP . The optimizations of C code, assembly language , storage space and the structure of interpolation algorithm are used to the decoder, and the results of every method of optimization have been got. As to QCIF the final result is three times as fast as the repotted one. It has met the real time need.
引文
[1] 数字音视频编解码技术标准工作组. 先进音视频编码(第7部分:移动视频)[s],2005.
[2] 李虓江, 陈抗生.一种H.264视频流自适应率失真优化编码算法[J].电路与系统学报,2005(6).
[3] 郑庆辉,何加铭,范志蔚等.一种H.264自适应块搜索范围快速运动估计算法[J].中国图象图形学报,2005,(11).
[4] Hyiingjoon Kim,Yucel Alriinhosnk. RATE-DISTORTION OPTIMAL JOINT MACROBLOCK MODE SELECTION AND MOTION ESTIMATION FOR MPEG-LIKE VIDEO CODERS[J]. 2004 International Conference on Image Processing (ICIP),2004:1137-1140.
[5]朱红,吴成柯,方勇.h.264宏块模式的一种快速判决方案[J].电子学报,2005(6).
[6] 王正宁,彭强,诸昌钤,.一种基于残差分析的H.264/AVC帧间块尺寸选择快速预测算法.中国图象图形学报,2005,(12)
[7] 陈梅芳. 基于TMS320DM642的H.264解码器优化[J].现代电子技术,2006(3)
[8] 何莉莉.基于TMS320DM642的H.264视频编码器设计[J].西华大学学报,2005(5)
[9] 张彤宇,苏睿, 刘宝兰.H.264/AVC视频编码器在DM642平台上的实现与优化[J].微电子学与计算机,2005(12).
[10]侯志强,韩崇昭,郑林等. 一种基于视觉模型的边缘检测阈值选择策略[J].光电工程,2004(2):59-62.
[11]Texas Instruments.TMS320C6000 Optimizing Compiler User’s Guide,2004(5)
[12] Texas Instruments. TMS320C6000 CPU and Instruction Set Reference Guide,2000(10)
[13] Muhammad Owais Khan,Umar Khan,Sardar Adnan Rahim.Optimization of Motion Compensation for H.264 Decoder by Pre-Calculation[J]Multitopic Conference, 2004. Proceedings of INMIC 2004. 8th International
[14] Texas Instruments.TMS320C6000 Programmer’s Guide,2002(10)
[15] Texas Instruments.TMS320C6000 CPU and Instruction Set Reference Guide,2000(10)
[16] Texas Instruments.TMS320C6000 DSP Enhanced Direct Memory Access (EDMA) Controller Reference Guide,2004(11)
[17] Texas Instruments.TMS320C6000 DSP External Memory Interface (EMIF) Reference Guide,2004(4)
[18] P.K. Merakos, K. Masselos, T. Stouraitis and C.E. Goutis. Optimisation techniques for reducing global bus switching activity in reialisations of sum-of-products computations in DSP systems[J]. IEE Proc Ciruirr Devices syst,2003(2)
[19] W. Lin, B. Tye, E. Ong, C. Xiong.Systematic Analysis and Methodology of Real-time DSP Implementation for Hybrid Video Coding[J].IEEE,1999
[20] 段大高,崔岩松,邓中亮.H.264的帧间宏块模式选择算法[J].现代有线传输
[21] 季虎,孙即祥,邵晓芳等.图像边缘提取方法及展望[J]. 计算机工程与应用,2004
[22] T. Ha, S. Lee. J. Kim.Motion compensated frame rate conversion by overlapped block-based motion estimation algorithm.2004 IEEE International Symposium on Consumer Electronics – Proceedings,2004:345～350
[23] 胡金波,陈慧剑.DSP 环境下 C 语言编程的优化实现[J].现代电子技术,2002(12):82～84
[24] Thomas Wiegand,Michael Lightstone et,Rate-Distortion Optimizated Mode se1ection for Very Low Bit Video Coding and the Emerging H.263 Standard,IEEE Trans. Circuits and Systems for Video Technology,1996(4)
[25] Yao Wang Ya-Qin Zhang 著,侯正信、杨喜、王文全等译,视频处理与通信,电子工业出版社,2003
[26] 吴乐南,数据压缩,南京:东南大学出版社,2000
[27] 安维嵘,张旭东.H.264视频解码器在C6416 DSP上的实现.电子技术应用,2004(09):46~48
[28] R. Zhang, S. L. Regunathan, K. Rose.Video Coding with Optimal Inter/Intra-Mode Switching for Packet Loss Resilience.IEEE Journal on Selected Areas in Communications, 2000,6,966~976
[29]Anurag Jain, Ratna Reddy, Jeremiah Golston, Jagadeesh Sankaran,Programmable Real-time MPEG-2 Encoding, GSPx 2002.
[30] Ismaeil, I.R.; Docef, A., Kossentini, F., Ward, R. Motion estimation using long-term motion vector prediction. Data Compression Conference, Proc. 1999 Page(s): 531.
[31] 钟玉涿,王琪,贺玉文. 基于对象的多媒体数据压缩编码国际标准——MPEG-4. 北京:科学出版社. 2000
[32] Thomas Wiegand, Bernd Girod. Multi-Frame Motion-Compensated Prediction for Video Transmission. America: Kluwer Academic Pulishers. 2001
[33] Wiegand T, Zhang X Z, Girod B. Long-Term Memory Motion-Compensated Prediction.IEEE Transaction on Circuits and Systems for Video Technology, 1999, 9(1):70-84
[34] Texas Instruments. TMS320C6000 Programmer’s Guide. TI Digital Signal Processing Solutions. 2000
[35] 安 维 嵘.H.264 解码算法的研究及其在 DSP 平台的实现[学位论文].清华大学
[36] 柳金峰.基于 DSP 的 AVS-M 视频压缩算法的研究和实现[学位论文].吉林大学
[37]李杰.h.264 解码器的 DSP 移植及优化[学位论文].华侨大学
[2] 李虓江, 陈抗生.一种H.264视频流自适应率失真优化编码算法[J].电路与系统学报,2005(6).
[3] 郑庆辉,何加铭,范志蔚等.一种H.264自适应块搜索范围快速运动估计算法[J].中国图象图形学报,2005,(11).
[4] Hyiingjoon Kim,Yucel Alriinhosnk. RATE-DISTORTION OPTIMAL JOINT MACROBLOCK MODE SELECTION AND MOTION ESTIMATION FOR MPEG-LIKE VIDEO CODERS[J]. 2004 International Conference on Image Processing (ICIP),2004:1137-1140.
[5]朱红,吴成柯,方勇.h.264宏块模式的一种快速判决方案[J].电子学报,2005(6).
[6] 王正宁,彭强,诸昌钤,.一种基于残差分析的H.264/AVC帧间块尺寸选择快速预测算法.中国图象图形学报,2005,(12)
[7] 陈梅芳. 基于TMS320DM642的H.264解码器优化[J].现代电子技术,2006(3)
[8] 何莉莉.基于TMS320DM642的H.264视频编码器设计[J].西华大学学报,2005(5)
[9] 张彤宇,苏睿, 刘宝兰.H.264/AVC视频编码器在DM642平台上的实现与优化[J].微电子学与计算机,2005(12).
[10]侯志强,韩崇昭,郑林等. 一种基于视觉模型的边缘检测阈值选择策略[J].光电工程,2004(2):59-62.
[11]Texas Instruments.TMS320C6000 Optimizing Compiler User’s Guide,2004(5)
[12] Texas Instruments. TMS320C6000 CPU and Instruction Set Reference Guide,2000(10)
[13] Muhammad Owais Khan,Umar Khan,Sardar Adnan Rahim.Optimization of Motion Compensation for H.264 Decoder by Pre-Calculation[J]Multitopic Conference, 2004. Proceedings of INMIC 2004. 8th International
[14] Texas Instruments.TMS320C6000 Programmer’s Guide,2002(10)
[15] Texas Instruments.TMS320C6000 CPU and Instruction Set Reference Guide,2000(10)
[16] Texas Instruments.TMS320C6000 DSP Enhanced Direct Memory Access (EDMA) Controller Reference Guide,2004(11)
[17] Texas Instruments.TMS320C6000 DSP External Memory Interface (EMIF) Reference Guide,2004(4)
[18] P.K. Merakos, K. Masselos, T. Stouraitis and C.E. Goutis. Optimisation techniques for reducing global bus switching activity in reialisations of sum-of-products computations in DSP systems[J]. IEE Proc Ciruirr Devices syst,2003(2)
[19] W. Lin, B. Tye, E. Ong, C. Xiong.Systematic Analysis and Methodology of Real-time DSP Implementation for Hybrid Video Coding[J].IEEE,1999
[20] 段大高,崔岩松,邓中亮.H.264的帧间宏块模式选择算法[J].现代有线传输
[21] 季虎,孙即祥,邵晓芳等.图像边缘提取方法及展望[J]. 计算机工程与应用,2004
[22] T. Ha, S. Lee. J. Kim.Motion compensated frame rate conversion by overlapped block-based motion estimation algorithm.2004 IEEE International Symposium on Consumer Electronics – Proceedings,2004:345～350
[23] 胡金波,陈慧剑.DSP 环境下 C 语言编程的优化实现[J].现代电子技术,2002(12):82～84
[24] Thomas Wiegand,Michael Lightstone et,Rate-Distortion Optimizated Mode se1ection for Very Low Bit Video Coding and the Emerging H.263 Standard,IEEE Trans. Circuits and Systems for Video Technology,1996(4)
[25] Yao Wang Ya-Qin Zhang 著,侯正信、杨喜、王文全等译,视频处理与通信,电子工业出版社,2003
[26] 吴乐南,数据压缩,南京:东南大学出版社,2000
[27] 安维嵘,张旭东.H.264视频解码器在C6416 DSP上的实现.电子技术应用,2004(09):46~48
[28] R. Zhang, S. L. Regunathan, K. Rose.Video Coding with Optimal Inter/Intra-Mode Switching for Packet Loss Resilience.IEEE Journal on Selected Areas in Communications, 2000,6,966~976
[29]Anurag Jain, Ratna Reddy, Jeremiah Golston, Jagadeesh Sankaran,Programmable Real-time MPEG-2 Encoding, GSPx 2002.
[30] Ismaeil, I.R.; Docef, A., Kossentini, F., Ward, R. Motion estimation using long-term motion vector prediction. Data Compression Conference, Proc. 1999 Page(s): 531.
[31] 钟玉涿,王琪,贺玉文. 基于对象的多媒体数据压缩编码国际标准——MPEG-4. 北京:科学出版社. 2000
[32] Thomas Wiegand, Bernd Girod. Multi-Frame Motion-Compensated Prediction for Video Transmission. America: Kluwer Academic Pulishers. 2001
[33] Wiegand T, Zhang X Z, Girod B. Long-Term Memory Motion-Compensated Prediction.IEEE Transaction on Circuits and Systems for Video Technology, 1999, 9(1):70-84
[34] Texas Instruments. TMS320C6000 Programmer’s Guide. TI Digital Signal Processing Solutions. 2000
[35] 安 维 嵘.H.264 解码算法的研究及其在 DSP 平台的实现[学位论文].清华大学
[36] 柳金峰.基于 DSP 的 AVS-M 视频压缩算法的研究和实现[学位论文].吉林大学
[37]李杰.h.264 解码器的 DSP 移植及优化[学位论文].华侨大学