H.263视频编解码系统优化及实现
摘要
本文深入讨论和研究了低比特率通信系统中视音频信号的实时处理及优化技术，并进行了软件实现。其中视频编解码遵循ITU-T H．263协议，音频编解码基于ITU-T G．723协议。视音频数据使用RTP协议进行实时传输，并采用了误码掩盖技术。
文中首先对视音频信号的编解码方法进行了讨论，在对视音频压缩协议H．263和G．723进行分析和软件实现的基础上，根据编解码的特点对运动估计和DCT模块，利用MMx技术提出并实现了优化算法。并且针对块匹配运动估计给出了一种改进的钻石搜索算法代替原有的三步搜索法。实验结果证明，经算法改进和MMX技术优化的新钻石搜索法远胜于三步搜索法，大大提高了程序运行的效率，使系统达到了软件实时编解码的需要。然后研究了如何用软件的方法实现视音频数据的采集与播放，并讨论了视音频数据的同步问题。接着讨论了纠错编码，采用RS编码作为前向纠错码，提出了一种基于运动特征的自适应误码掩盖方法，并采用RTP协议进行视音频数据的实时传输及带宽控制，
本文的研究成果可以直接用于远程教育、视频会议、可视电话、视频监控等各种通信系统，具有较大的研究价值及广泛的实用价值。
This paper deeply discuss the video and audio real-time processing technology and its optimize technology of low bits rate communication systems. The system has been realized with software. Video compression are based on the H.263 algorithm, audio compression are based on the H.723 algorithm, data transmit by RS code and RTF protocol.
At first this paper discuss the method of audio and video signal encoder, optimize the motion estimate and DCT modules with MMX technology. A new diamond search algorithm for fast-matching motion estimation is proposed to replace the traditional three-step search algorithm in the motion estimation region. Experiments results demonstrate that the application on the new diamond search algorithm optimized by MMX has magnificently improved the efficiency of compression ratio and computing speed. Then it is dicussed that how to capture and play and synchronize the video and audio data with software. Error correcting code are also discussed, the RS code has been used to correcting code. It has put forward a kind of method of self-adaptation based on motion characteristic to correcting error code. Finally, based on the RTP protocol, real-time transmission of video and audio data is realized with software, and a bandwidth control algorithm based on RTP is introduced.
The research results of this paper can be used in the long-distance education, video conference, videophone, video real-time control, and various kinds of communication systems. It has greater research value and extensive practical value.
文中首先对视音频信号的编解码方法进行了讨论，在对视音频压缩协议H．263和G．723进行分析和软件实现的基础上，根据编解码的特点对运动估计和DCT模块，利用MMx技术提出并实现了优化算法。并且针对块匹配运动估计给出了一种改进的钻石搜索算法代替原有的三步搜索法。实验结果证明，经算法改进和MMX技术优化的新钻石搜索法远胜于三步搜索法，大大提高了程序运行的效率，使系统达到了软件实时编解码的需要。然后研究了如何用软件的方法实现视音频数据的采集与播放，并讨论了视音频数据的同步问题。接着讨论了纠错编码，采用RS编码作为前向纠错码，提出了一种基于运动特征的自适应误码掩盖方法，并采用RTP协议进行视音频数据的实时传输及带宽控制，
本文的研究成果可以直接用于远程教育、视频会议、可视电话、视频监控等各种通信系统，具有较大的研究价值及广泛的实用价值。
This paper deeply discuss the video and audio real-time processing technology and its optimize technology of low bits rate communication systems. The system has been realized with software. Video compression are based on the H.263 algorithm, audio compression are based on the H.723 algorithm, data transmit by RS code and RTF protocol.
At first this paper discuss the method of audio and video signal encoder, optimize the motion estimate and DCT modules with MMX technology. A new diamond search algorithm for fast-matching motion estimation is proposed to replace the traditional three-step search algorithm in the motion estimation region. Experiments results demonstrate that the application on the new diamond search algorithm optimized by MMX has magnificently improved the efficiency of compression ratio and computing speed. Then it is dicussed that how to capture and play and synchronize the video and audio data with software. Error correcting code are also discussed, the RS code has been used to correcting code. It has put forward a kind of method of self-adaptation based on motion characteristic to correcting error code. Finally, based on the RTP protocol, real-time transmission of video and audio data is realized with software, and a bandwidth control algorithm based on RTP is introduced.
The research results of this paper can be used in the long-distance education, video conference, videophone, video real-time control, and various kinds of communication systems. It has greater research value and extensive practical value.
引文
[1] DRAFT ITU-T Recommendation H. 263 (05/1996),Video Coding for Low Bit Rate Communication.
[2] CCITT Recommendation H.261 (1990), Video coding for audio visual services at p×64bits/s, CCITT study Group xv, Dec.
[3] ITU-T Recommendation H.263v2 ("H. 263+",02/1998),Video Coding for Low Bit Rate Communication.
[4] ITU-T Recommendation H. 263v3 ("H. 263++",11/2000),Video Coding For Low Bit Rate Communication.
[5] ITU-T Rec. H. 245 (07/2001),Control protocol for multimedia communication.
[6] ITU-T Recommendation H. 263 Appendix Ⅲ(06/2001),Examples for H. 263 Encoder/Decoder Implementations.
[7] RFC1889 "RTP: A Transport Protocol for Real-Time Applications"
[8] RFC1890 "RTP Profile for Audio and Video Conferences with Minimal Control"
[9] Y. Lin and S. Tai. Fast full-search block-matching algorithm for motion-compensated video compression, IEEE Transactions on Communications, vol. 45, no. 5, May 1997, pp. 527-531.
[10] R. Li, B.Zeng, and M.L. Liou, A new three-step search algorithm for block motion estimation. IEEE Trans. Circuits Syst. Video Technol.,VoL 4, pp. 438-442, Aug. 1994
[11] M. Ghanbari, The cross-search algorithm for motion estimation. IEEE Trans. Commun.,vol. 38, pp. 950-953, July 1990.
[12] S. Zhu and K.K. Ma, A new diamond search algorithm for fast block-matching motion estimation. IEEE Trans. Image Processing, vol. 9, pp. 287-290, Feb. 2000
[13] J.B. Xu, L.M. Po. and C.K. Cheung. Adaptive motion tracking block matching algorithms for video coding. IEEE Trans. Circuits Syst. Video Technol. Vol. 9, pp. 1025-1029,Oct. 1999
[14] Alice Yu, Motion Search Performance using the H. 263 Encoder, EE392c Project, December 1,1997
[15] Richard c．Leinecker，Tom Archer著，张艳，王文学，张谦等译．Visual c++6 宝典，北京：电子工业出版社，2001．
[16] 周璇、谭径微、余松煜．H．263中预先判别全零系数的新方法．上海交通大学学报，Vol．32 No．9，Sep．1998，107～109
[17] 吴乐南．数据压缩．电子工业出版社，2000
[18] 余松煜、郑志航．图像通信中的运动估计．通信学报，Vol．14 No．2，March
1993
[19] 苏育挺，张春田．极低码率视频编码中运动估值技术的研究．通信学报，Vol．20 No．6，June 1999，8~14
[20] 陈晓、朱耀庭、朱光喜。一种新的快速H．263算法的研究．华中理工大学学报，Vol．27 No．3，Mar．1999，65～68
[21] 胡师彦，基于视频监控应用的改进的H．263算法，南昌大学学报(工科版)，Vol．24 No．1，Mar．2002，44～47
[22] Anthony．]ones、Jim Ohlund编著《Windows 网络编程技术》机械工业出版社2000
[23] 罗明辉、朱秀昌、王厚大．基于．H．263视频解码的误码掩盖技术研究．通信学报，Vol．20 No．5，May 1999，94～96
[24] 梁柱．H．263编码视频流的混合错误掩盖．电视技术，2002(2)，72～74
[25] 纪其进、董育宁．H．263的差错恢复视频编码及其在IP网络传输中的应用．电视技术，2002(1)，63～66
[26] 胡广书．数字信号处理——理论、算法与实现．清华大学出版社，2000
[27] 胡琳蓉、朱秀昌．一种适用于H．263的运动估计搜索算法，通信学报，Vol．21 No．6，June 2000，65～69
[28] 余松煜，张文军，孙军编著 《现代图像信息压缩技术》科学出版社1999
[29] 余松煜，郑志航编著《图像通信中的运动估值》通信学报 1993
[30] Intel公司编著李晖译《Intel体系结构MMX~TM技术开发指南》电子工业出版社1997
[30] 黎洪松编著《数字视频技术及其应用》清华大学出版社 1997
[32] 杨行峻、迟惠生等编著《语音信号数字处理》电了工业出版社1995
[33] 易克初等编著《语音信号处理》国防工业出版社2000
[34] 蔡皖东编著《多媒体通信技术》西安电子科技大学出版社2000
[35] 李小平、曲大成编著《多媒体网络通信》北京理工大学出版社2001
[36] 高文编著《多媒体数据压缩技术》电子工业出版社1994
[37] Peter Aitken Scott Jarol编著李鹤文张文新译《Visual C++多媒体开发指南》科学出版社1996
[38] 张云、林钟编著《PC平台新技术MMX》东南大学出版社1998
[39] 吴成柯、戴善荣、陆心如编著《图像通信》西安电子科大出版社2000
[40] 罗明宇、陶孜谨、卢锡城，《RTP在网络视频传输中的实现研究》，计算机工程，2000年第9期
[2] CCITT Recommendation H.261 (1990), Video coding for audio visual services at p×64bits/s, CCITT study Group xv, Dec.
[3] ITU-T Recommendation H.263v2 ("H. 263+",02/1998),Video Coding for Low Bit Rate Communication.
[4] ITU-T Recommendation H. 263v3 ("H. 263++",11/2000),Video Coding For Low Bit Rate Communication.
[5] ITU-T Rec. H. 245 (07/2001),Control protocol for multimedia communication.
[6] ITU-T Recommendation H. 263 Appendix Ⅲ(06/2001),Examples for H. 263 Encoder/Decoder Implementations.
[7] RFC1889 "RTP: A Transport Protocol for Real-Time Applications"
[8] RFC1890 "RTP Profile for Audio and Video Conferences with Minimal Control"
[9] Y. Lin and S. Tai. Fast full-search block-matching algorithm for motion-compensated video compression, IEEE Transactions on Communications, vol. 45, no. 5, May 1997, pp. 527-531.
[10] R. Li, B.Zeng, and M.L. Liou, A new three-step search algorithm for block motion estimation. IEEE Trans. Circuits Syst. Video Technol.,VoL 4, pp. 438-442, Aug. 1994
[11] M. Ghanbari, The cross-search algorithm for motion estimation. IEEE Trans. Commun.,vol. 38, pp. 950-953, July 1990.
[12] S. Zhu and K.K. Ma, A new diamond search algorithm for fast block-matching motion estimation. IEEE Trans. Image Processing, vol. 9, pp. 287-290, Feb. 2000
[13] J.B. Xu, L.M. Po. and C.K. Cheung. Adaptive motion tracking block matching algorithms for video coding. IEEE Trans. Circuits Syst. Video Technol. Vol. 9, pp. 1025-1029,Oct. 1999
[14] Alice Yu, Motion Search Performance using the H. 263 Encoder, EE392c Project, December 1,1997
[15] Richard c．Leinecker，Tom Archer著，张艳，王文学，张谦等译．Visual c++6 宝典，北京：电子工业出版社，2001．
[16] 周璇、谭径微、余松煜．H．263中预先判别全零系数的新方法．上海交通大学学报，Vol．32 No．9，Sep．1998，107～109
[17] 吴乐南．数据压缩．电子工业出版社，2000
[18] 余松煜、郑志航．图像通信中的运动估计．通信学报，Vol．14 No．2，March
1993
[19] 苏育挺，张春田．极低码率视频编码中运动估值技术的研究．通信学报，Vol．20 No．6，June 1999，8~14
[20] 陈晓、朱耀庭、朱光喜。一种新的快速H．263算法的研究．华中理工大学学报，Vol．27 No．3，Mar．1999，65～68
[21] 胡师彦，基于视频监控应用的改进的H．263算法，南昌大学学报(工科版)，Vol．24 No．1，Mar．2002，44～47
[22] Anthony．]ones、Jim Ohlund编著《Windows 网络编程技术》机械工业出版社2000
[23] 罗明辉、朱秀昌、王厚大．基于．H．263视频解码的误码掩盖技术研究．通信学报，Vol．20 No．5，May 1999，94～96
[24] 梁柱．H．263编码视频流的混合错误掩盖．电视技术，2002(2)，72～74
[25] 纪其进、董育宁．H．263的差错恢复视频编码及其在IP网络传输中的应用．电视技术，2002(1)，63～66
[26] 胡广书．数字信号处理——理论、算法与实现．清华大学出版社，2000
[27] 胡琳蓉、朱秀昌．一种适用于H．263的运动估计搜索算法，通信学报，Vol．21 No．6，June 2000，65～69
[28] 余松煜，张文军，孙军编著 《现代图像信息压缩技术》科学出版社1999
[29] 余松煜，郑志航编著《图像通信中的运动估值》通信学报 1993
[30] Intel公司编著李晖译《Intel体系结构MMX~TM技术开发指南》电子工业出版社1997
[30] 黎洪松编著《数字视频技术及其应用》清华大学出版社 1997
[32] 杨行峻、迟惠生等编著《语音信号数字处理》电了工业出版社1995
[33] 易克初等编著《语音信号处理》国防工业出版社2000
[34] 蔡皖东编著《多媒体通信技术》西安电子科技大学出版社2000
[35] 李小平、曲大成编著《多媒体网络通信》北京理工大学出版社2001
[36] 高文编著《多媒体数据压缩技术》电子工业出版社1994
[37] Peter Aitken Scott Jarol编著李鹤文张文新译《Visual C++多媒体开发指南》科学出版社1996
[38] 张云、林钟编著《PC平台新技术MMX》东南大学出版社1998
[39] 吴成柯、戴善荣、陆心如编著《图像通信》西安电子科大出版社2000
[40] 罗明宇、陶孜谨、卢锡城，《RTP在网络视频传输中的实现研究》，计算机工程，2000年第9期