基于MMX技术的H.26L编码速度的研究与改进
详细信息    本馆镜像全文|  推荐本文 |  |   获取CNKI官网全文
摘要
随着数字技术和通讯技术的发展,图像视频编码技术和编码标准也得到了长足的进步。H.26L便是目前面向无线通信和多播的低码率视频传输标准的最新版本。该标准具有压缩率高、差错恢复能力强、适用范围广等优点。为达到理想的数据压缩率,H.26L编码器采用了很多新的编码算法,这些算法极大地增加了编码器的计算复杂性,使其目前在PC环境下的实时应用受到很多限制。在要加速H.26L的实时编码速度,一个行之有效的方法便是采用MMX技术对其执行时间最多、执行次数最多的部分代码进行优化,以减少编码时间。本文以H.26L的测试模型TML8.0为实验平台,对其编码器进行了详细的分析和性能测试,找出最值得优化的程序代码,并在此基础上采用MMX技术对其进行优化,加快编码速度。本论文主要包含以下内容:
     1、简要介绍了视频编码的基本方法和H.26L编码标准的主要特点和新引进的算法;并对MMX技术相关知识进行介绍,说明MMX技术进行优化的步骤。这部分知识是进行本论文研究的基础。
     2、通过对H.26L视频编码的基本原理以及H.26L编码器程序TML8.0的分析及测试,找出H.26L视频编码中耗时最多、执行次数最多的模块。
     3、对以上找出的模块进行MMX优化,并对其试验结果进行了分析和对比。通过从对整像素搜索的优化效果的分析,得出不必对1/2像素和1/4像素搜索进行优化的原因。
     分析和试验表明,通过MMX优化的方式可以很大程度的提高H.26L视频编码的处理速度,使得该标准可以更好的适用于PC机环境下的多种不同的视频应用。
With the development of digital technique and communication technique, multimedia video coding technique and corresponding coding standard are developing accordingly.
    H.26L, a video-coding standard proposed by ITU, is the latest standard for video communication and multicast in low bit rate conditions. H.26L has such advantages as high compression rate, strong error resilience capacity, high adaptability and so on. In order to gain good coding result, H.26L adopts so many new algorithms that the amount of computing increase greatly. As a result, its applications in PC environment are restricted. In order to accelerate H.26L real-time coding, MMX technique might be effectively used to optimize part code that takes the most time and executes the most times, which may help to decrease the coding time.
    Based on the test model long-term 8.0(TML8.0) of H.26L and through a detailed analysis and capability test of the encoder, the present paper finds out the code that is worthy of optimizing and it also finds out the way to optimize the code and accelerate coding by using MMX technique. The main contents of the paper are as follows:
    1. The thesis gives a brief introduction to basic ways of video-coding, the main features of H.26L vide encoder, primary theories, and the correlative knowledge of MMX technique. The optimize step of MMX technique was introduced as foundations of this paper.
    2. The paper points out that the module that takes the most time and executes the most times can be found out through the analysis of the basic principle of H.26L Video Coding and coding procedure TML8. Oof H.26L coder.
    3.The present paper optimizes the above-mentioned module by using MMX technique, analyzes and contrasts the result of examination. We can see the reason why we need not optimize the half pixel search and quarter pixel search by analyzing the optimizing effect of integer pixel search.
    The analysis and simulation results show that MMX technique can greatly speed up the processing of H.26L encoder and make it possible for the H.26L to be used in PC environment in many video applications.
引文
[1]ITU-T/SG16/VCEG (Q. 6),H.26L Test Model Long-Term Number 8 (TML-8).http://standards.pictel.com/ftp/video-site/O109_San/VCEG-N10.doc
    [2]Video coding for Low Bitrate Communication,ITU-T Recommendation H.263 Version 2,1998
    [3]Cté,G.,and F.Kossentini,"Optimal Intra Coding of Macroblocks for Robust{H.263} Video Communication over the Internet",submitted to Special Issue of Image Communication on "Real-time Video over the Internet" EUROSIP Visual Communication,Sept.1998.
    [4]Yao Wang, Gin-fan Zhu,"Error Control and Concealment for Video Communication:A Review,"Proc.IEEE,vol.86,May 1998
    [5]Jerry D.Gibson,Toby Berger,Tom Lookabaugh等著.多媒体数字压缩原理与标准.李煜晖,朱山风,段上为等译.北京:电子工业出版社,2000年8月
    [6]张明德,王永东.视频会议系统原理与应用.北京:北京希望电脑公司,1999年8月
    [7]吴乐南.数据压缩.北京:电子工业出版社,2000年6月
    [8]Ralf Steinmetz,Klara Nanrstedt著.多媒体技术:计算、通信和应用.潘志庚、叶绿、耿卫东等译.北京:清华大学出版社,2000年12月
    [9]黎洪松.数字视频技术及其应用.北京:清华大学出版社,1997年5月
    [10]钟玉琢.多媒体技术(高级).北京:清华大学出版社,1999年7月
    [11]TEKALP,A.MURAT著.数字视频处理,崔之祜译.北京:电子工业出版社,1998年7月
    [12]胡国荣.数字视频压缩及其标准.北京:北京广播学院出版社,1999年12月
    [13]沈兰荪.视频编码与低速率传输.北京:电子工业出版社,2001年
    [14]杨学良,张占军.分布式多媒体计算机系统教程.北京:电子工业出版社,2002年
    [15]李立杰.多媒体及其通信技术.北京:机械工业出版社,2002年
    [16]欧建平,娄生强.网络与多媒体通信技术.北京:人民邮电出版社,2002年
    [17]李在铭.数字图像处理、压缩与识别技术.成都:电子科技大学出版
    
    社,2000年11月
    [18]沈兰荪.图像编码与异步传输.北京:人民邮电出版社,1998年5月
    [19]余松煜.现代图像信息压缩技术.北京:科学出版社,1998年11月
    [20]郝杰.实用多媒体技术及其C语言实现.北京:电子工业出版社,1995年11月
    [21]彭强,余滨.多媒体个人计算机实用技术.成都:西南交通大学出版社,1996年10月
    [22]吴乐南.PC平台新技术MMX——开发编程指南.南京:东南大学出版社,1997年10月
    [23]张云.PC平台新技术MMX——应用编程实例.南京:东南大学出版社,1998年5月
    [24]骆立俊,邹采荣,何振亚,一种新的视频编码的块运动估计算法,通信学报,Vol.21,No.2,February,2000:pp55-59
    [25]Anthony Joch, Faouzi Kossentini and Panos Nasiopoulos. A Performance Analysis of the ITU-T Draft H. 26L Video Coding Standard. http://pv2002.ece.cmu.edu/papers/35-nteueubjoc.pdf
    [26]Qinqing Zhang and Saleem A. Kassam, Hybrid ARQ with Selective Combining for Video Transmission over Wireless Channels, IEEE International Conference on Image Processing, Vol. 2, 1997:pp. 692-695
    [27]C.C. Liao, T.H. Lee, and P.C. Chang, Error Concealment with Macroblock Interleaving for H. 263 Video Transmission, in Proc. IPPR Conf Computer Vision, Graphics, and Image Processing, CVGIP'97, TaiChung, Taiwan, Aug. 1997 :pp570-577
    [28]J. Villasennor, Y-Q. Zhang, J. Wen, "Robust Video Coding Algorithms and System," Proc. IEEE, vol. 87, Oct. 1999: pp. 1724-1733
    [29]J. Wen and J. D. Villasenor, "A class of reversible variable length codes for robust image and video coding," in Proc. 1997 IEEE Int. Conf. Image Processing (ICIP), vol. 2, Santa Barbara, CA, Oct. 1997: pp. 65-68.
    [30]骆立俊,邹采荣,何振亚,一种新的视频编码的块运动估计算法,通信学报,Vol.21,No.2,February,2000:pp55-59
    [31]G. Cote and F. Kossentini, "Optimal intra coding of blocks for robust video communication over the internet," in Image Commun., Sept.
    
    1999, pp. 25-34
    [34]G.J. Sullivan and T. Wiegand, "Rate-distortion optimization for video compression," IEEE Signal Proc. Magazine, Vol. 15, No. 6, Nov. 1998: pp. 23-50
    [35]Rui Zhan, Shankar L. Regunatha and Kenneth Ros, Video Coding with Optimal Inter/Intra-ModeSwitching for Packet Loss Resilience, IEEE JOURNAL ON SELECTED AREAS IN COMMUNICATIONS, VOL. 18, NO. 6, JUNE 2000
    [36]颜金尧,宋宜纯,低码率视频压缩技术地块效应消除办法,通信学报,Vol.21,No.4,April,2000:pp42-47
    [37]M. Wada, Selective Recovery of Video Packet Loss Using Error Concealment", IEEE Journal on Selection Areas of Communication, June 1989:pp. 807-814
    [38]Michele Zorzi, Performance of FEC and ARQ Error Control in Bursty Channels under Delay Constrants, Vtc' 98, Ottawa, Canada, May 1998
    [39]胡国荣,数字视频压缩及其标准,北京:北京广播学院出版社,1999
    [40]钟玉琢.MPEG-2运动图像压缩编码国际标准及MPEG的新进展.北京:清华大学出版社,2002年
    [41]钟玉琢.基于对象的多媒体数据压缩编码国际标准:MPEG-4及其校验模型.北京:科学出版社,2000年
    [42]Intel Corporation, Pentium Processor Family Developer's Manual (Volume 1,2and 3). http://www.intel.com
    [43]Intel Corporation, Intel Architecture MMX Technology Programmer's Reference Manual. March 1996
    [44]Intel Corporation, Intel Architecture MMX Technology Developer' Reference Manual. March 1996
    [45]Intel Corporation, Use MMX Instruction to Compute the Diffence in Motion Estimation. March 1996
    [46]Intel Corporation, Use MMX Instruction in a Fast IDCT Algorithm for MPEG Decoding. March 1996
    [47]Vancouver, British Columbia, Canada V6J 1P2. Emerging H. 26L Standard: Overview and TMS320C64x Digital Media Platform Implementation. http://www.ubvideo.com/public/h261-white_paper.pdf
    
    
    [48]Mrudula Yadav and Gayathri Venkat. Modeling and Simulation of H. 26L Encoder. http://www.ece.utexas.edu/~bevans/courses/ee382c/projects/spring02/venkat-yadav/FinalReport.pdf
    [49]Antti Hallapuro, Ville Lappalainen and Time D. Hamallainen. Performance analysis of low bit rate H. 26L video encoder. http://pv2002.ece.cmu.edu/papers/35-nteueubjoc.pdf
    [50]彭强,视频通信差错恢复技术,计算机应用研究,7期,pp9-12,July,2001
    [51]丁志强,MMX技术在视频压缩中的应用,数字通讯,2000,3:51-53
    [52]毕厚杰,多媒体信息的传输与处理,北京:人民邮电出版社,1999
    [53]朱秀昌,多媒体网络通讯技术与应用,北京:电子工业出版社,1998