网络视频会议系统的设计与实现
摘要
网络视频会议系统是多个与会者通过网络传输图像、声音、文字等信息，来实现异地会议的一种通信系统。目前的视频会议系统大多是基于H.323规范的集中式会议系统，由于其包括了多点控制单元MCU、多点处理器MP、网关以及网守等设备，尽管性能较优，但其价格昂贵，配置要求较高，不易普遍推广。
本文提出并实现了一种低成本高质量的新型视频会议系统，其服务器端能任意指定n个与会者，各客户端实时捕获与会者的视音频数据并将之传送到服务器端，服务器端合并处理各路视音频数据后，以组播的方式回传给各与会者客户端，同时推发布到流媒体服务器上，使得与会者之外的其他用户可通过Windows Media Player实时地观看到视频会议。本系统是由纯软件实现，数据回传给与会者时利用了分布式网络的组播功能，推发布给流媒体服务器时使用了流媒体传输协议，克服了基于硬件的网络视频会议系统所需要的特殊硬件以及固定通道的缺陷，大大降低了网络视频会议的成本。
本系统采用面向对象的编程方法，基于Windows平台，使用VC在DirectShow框架下编程并结合Windows Media技术实现视频会议。研究的内容有：根据TCP、UDP和RTP协议的特点，在网络传输的各个部分合理地选择传输协议，保证媒体数据的传输流畅性和响应实时性；采用抠像技术，将人物目标从背景中剥离；服务器端合并处理视音频数据，并以MPEG-4和ADPCM技术压缩视音频数据后，采用IP组播技术进行回传，节约网络带宽，优化网络传输；将合成后的数据以流的方式推发布到流媒体服务器上，使得非与会者能实时地观看视频会议。为实现各类之间的实时消息通信，本系统还设计了一个消息栈基类CMsgStation，其他业务逻辑类均继承它，这样某个类发送的消息可实时地被其它类所接收。由于采用了COM编程规范，极大地方便了软件升级和扩展，为后续工作的开展提供了基础。
Network video conference system is the communication system, where participants of the conference can transfer video, audio and words through Internet. At the present, most of video conference systems are the centralized systems based on H.323, which require some devices such as MCU, MP, Gateway and Gatekeeper. Owing to the expensive price and the higher requirement for the config, they are popularized difficultily despite the better performance.
A new low-cost and high-quality network video conference system is given and realized in the paper. N participants can be set in the server. Then the video and audio which are captured in the client will be transferred to the server. After the video and audio have been combined in the server, they will be returned to participants by multicast. Contemporarily, these combined data in the server will be pushed to the stream media server, which make other users watch the conference on real time by Windows Media Player. In the system, The combined data are sent to participants by Multicast and pushed to the stream media server by the stream media transfer protocol, which conquers the bugs where some specific hardware and fixed routeway should be provided by the network video conference based on hardware. As a result, the cost of the network video conference reduces enormously.
Following the OOP, we implement the network video conference system adopting VC, DirectShow and Windows Media technology on windows platform, where the researched contents are: According to the characteristic of TCP, UDP and RTP, we reasonably choose the protocol for every phase of data transfer in the network, insure the fluency of transferring media data and the real time of response; Using the knock out technology, we peel off the object from the background; After the video and audio have been combined and respectively compressed by MPEG-4 and ADPCM technology, they are fed back by IP multicast technology, which saves the network bandwidth and optimizes the network transfer; After the combined data in the server are pushed to the stream media server, other users can also watch the conference on real time. In order to communicate among all classes, we design a basic message stack class named CMsgStation, which will be inherited by other business logic classes. Consequently, the messages sent by a class can be received immediately by other classes. The system based on COM in the paper can be upgraded and extended conveniently.
本文提出并实现了一种低成本高质量的新型视频会议系统，其服务器端能任意指定n个与会者，各客户端实时捕获与会者的视音频数据并将之传送到服务器端，服务器端合并处理各路视音频数据后，以组播的方式回传给各与会者客户端，同时推发布到流媒体服务器上，使得与会者之外的其他用户可通过Windows Media Player实时地观看到视频会议。本系统是由纯软件实现，数据回传给与会者时利用了分布式网络的组播功能，推发布给流媒体服务器时使用了流媒体传输协议，克服了基于硬件的网络视频会议系统所需要的特殊硬件以及固定通道的缺陷，大大降低了网络视频会议的成本。
本系统采用面向对象的编程方法，基于Windows平台，使用VC在DirectShow框架下编程并结合Windows Media技术实现视频会议。研究的内容有：根据TCP、UDP和RTP协议的特点，在网络传输的各个部分合理地选择传输协议，保证媒体数据的传输流畅性和响应实时性；采用抠像技术，将人物目标从背景中剥离；服务器端合并处理视音频数据，并以MPEG-4和ADPCM技术压缩视音频数据后，采用IP组播技术进行回传，节约网络带宽，优化网络传输；将合成后的数据以流的方式推发布到流媒体服务器上，使得非与会者能实时地观看视频会议。为实现各类之间的实时消息通信，本系统还设计了一个消息栈基类CMsgStation，其他业务逻辑类均继承它，这样某个类发送的消息可实时地被其它类所接收。由于采用了COM编程规范，极大地方便了软件升级和扩展，为后续工作的开展提供了基础。
Network video conference system is the communication system, where participants of the conference can transfer video, audio and words through Internet. At the present, most of video conference systems are the centralized systems based on H.323, which require some devices such as MCU, MP, Gateway and Gatekeeper. Owing to the expensive price and the higher requirement for the config, they are popularized difficultily despite the better performance.
A new low-cost and high-quality network video conference system is given and realized in the paper. N participants can be set in the server. Then the video and audio which are captured in the client will be transferred to the server. After the video and audio have been combined in the server, they will be returned to participants by multicast. Contemporarily, these combined data in the server will be pushed to the stream media server, which make other users watch the conference on real time by Windows Media Player. In the system, The combined data are sent to participants by Multicast and pushed to the stream media server by the stream media transfer protocol, which conquers the bugs where some specific hardware and fixed routeway should be provided by the network video conference based on hardware. As a result, the cost of the network video conference reduces enormously.
Following the OOP, we implement the network video conference system adopting VC, DirectShow and Windows Media technology on windows platform, where the researched contents are: According to the characteristic of TCP, UDP and RTP, we reasonably choose the protocol for every phase of data transfer in the network, insure the fluency of transferring media data and the real time of response; Using the knock out technology, we peel off the object from the background; After the video and audio have been combined and respectively compressed by MPEG-4 and ADPCM technology, they are fed back by IP multicast technology, which saves the network bandwidth and optimizes the network transfer; After the combined data in the server are pushed to the stream media server, other users can also watch the conference on real time. In order to communicate among all classes, we design a basic message stack class named CMsgStation, which will be inherited by other business logic classes. Consequently, the messages sent by a class can be received immediately by other classes. The system based on COM in the paper can be upgraded and extended conveniently.
引文
[1] 张光华．视频会议系统体系标准及其QoS保证[J]．现代通信，2005(3)．
[2] 孟祥清，王建民．软件视频会议的实现[J]．视听界广播电视技术，2006(4)．
[3] 詹承华．视频会议系统中多点显示单元的设计与实现[D]．陕西：西北工业大学软件工程系，2006．
[4] 涂友华．基于SIP的IP视频会议系统的研究[D]．北京：北京邮电大学计算机应用系，2006．
[5] 吴震．视频会议系统的研究与实现[D]．四川：电子科技大学软件工程系，2006．
[6] 史臻赜．一种新型的通用视频会议系统研究与设计[J]．计算机与信息技术，2005(10)．
[7] 张新盛，郑建彬．基于软交换的IP视频会议技术研究[J]．计算机与信息技术，2005(10)．
[8] 陆其明．DirectShow实务精选[M]．北京：北京科海电子出版社，2004．
[9] 陆其明．DirectShow开发指南[M]．北京：清华大学出版社，2005．
[10] 杨淑莹．图像模式识别—VC++技术实现[M]．北京：清华大学出版社，北京交通大学出版社，2005．
[11] 杨明川．层次化的宽带流媒体服务质量保证[J]．信息网络，2005(2)．
[12] 徐瑛．DirectShow技术在远程教学系统中的应用研究[J]．职业技术学院学报，2006(5)．
[13] 顾学飞．基于DirectShow的广域网P2P视频会话系统[J]．江南大学学报，2006(5)．
[14] 丰伟伟．DirectShow播放视频和MP3[J]．电脑编程技巧与维护，2006(12)．
[15] Cline, L.S.; Du, J.; Keany, B.; Lakshman, K.; Maciocco, C.; Putzolu, D.M. DirectShow~(TM) RTP support for adaptivity in networked multimedia applications[J]. Multimedia Computing and Systems, 1998. Proceedings. IEEE International Conference. 1998(7).
[16] Microsoft Corporation. DirectShow 9.0 Document[Z]. Microsoft, 2003.
[17] Microsoft Corporation. Microsoft MSDN library[Z], http://www.microsoft.com.
[18] 刘平，高胜法．一种基于DirectShow COM接口的多媒体流解码程序的设计[J]．山东科学，2002(5)．
[19] 王厚雪．基于目标识别的视音频合成传输系统设计[D]．天津：天津理工大学计算机系，2005．
[20] Viriyaphol, p.; Jittawiriyanukoon, C. Efficient routing with clue protocol for IP based networks[J]. Mobile Technology, Applications and Systems, 2005 2nd International Conference, 2005(10).
[21] Sengupta, S.; Kumar, V.; Saha, D. Switched optical backbone for cost-effective scalable core IP networks[J]. Communications Magazine, IEEE, 2003(6).
[22] Young-ho Lee; Ki-Won Kwon; Jin-Tea Kim; Chul-Dong Lee. IP development and management of IP DB enabling efficient system-on-chip design[J]. ASICs, 1999. AP-ASIC'99. The First IEEE Asia Pacific Conference, 1999(8).
[23] 鲁宏伟，龚雄斌，孔华锋．基于接收端应用的TCP流量控制策略的研究[J]．小型微型计算机系统，2006(12)．
[24] 苑春苗，杨清永，王闵．IP网上稳定可靠的视频流系统设计与实现[J]．计算机工程，2003(7)．
[25] 余萍．基于Internet的实时视频传输技术浅析[J]．电力系统通信，2002(4)．
[26] Kai Xu; Ye Tian; Ansari, N. TCP-Jersey for wireless IP communications[J]. Selected Areas in Communications, IEEE Journal, 2004(5).
[27] Ebrahimi-Taghizadeh, S.; Helmy, A.; Gupta, S. TCP vs. TCP: a systematic study of adverse impact of short-lived TCP flows on long-lived TCP flows[J]. INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies. Proceedings IEEE, 2005(3).
[28] Girish Motwani; Gopinath, K. Evaluation of advanced TCP stacks in the iSCSI environment using simulation model[J]. Mass Storage Systems and Technologies, 2005. proceedings. 22nd IEEE/13th NASA Goddard Conference, 2005(3).
[29] 刘捷，龚根华．基于UDP的新型数据传输协议[J]．南昌大学学报(理科版)，2006(5)．
[30] 赵飞，叶震．UDP协议与TCP协议的对比分析与可靠性改进[J]．计算机技术与发展，2006(9)．
[31] Lam, P. P. -K.; Liew, S.C. UDP-Liter: an improved UDP protocol for real-time multimedia applications over wireless links[J]. Wireless Communication Systems, 2004. 1st International Symposium, 2004(9).
[32] Chang, R. K. C.; Fung, K. P. Transport layer proxy for stateful UDP packet filtering[J]. Computers and Communications, 2002. Proceedings. ISCC 2002. Seventh Internationsl Symposium, 2002(7).
[33] Mertz, F.; Engelke, U.; Vary, P.; Taddei, H.; Varga, I. Applicability of UDP-lite for voice over IP in UMTS networks[J]. Personal, Indoor and Mobile Radio Communications, 2005. PIMRC 2005. IEEE 16th Internatinal Symposium, 2005(9).
[34] 郑力明，张会汀．基于IP组播技术的分布式视频会议系统的设计与实现[J]．计算机工程与应用，2003(2)．
[35] 杨晓东，乐正友．IP组播技术及流式媒体在网上音视频广播中的应用[J]．现代电视技术，2002(4)．
[36] 王斌，刘增基．基于ACK和NACK的可靠组播传输协议的性能分析和比较[J]．电子学报，2001 (10)．
[37] Ki-ll Kim; Jeoung-Lak Ha; Eun-Hee Hyun; Sang-Ha Kim. Internet multicast provisioning issues for hierarchical architecture[J]. Networks, 2001. Proceedings. Ninth IEEE Internatinal Conference, 2001(10).
[38] Xue-Bin Liang. Matrix games in the multicast networks: maximum information flows with network switching[J]. Information Theory, IEEE Transactions, 2006(6).
[39] 王正勇，李永合．MPEG-4基于内容的视频编码研究[J]．四川大学学报(自然科学版)，2001(10)．
[40] 雷刚勇，解梅．基于DirectShow的MPEG-4远程视频监控系统[J]．视频技术应用与工程，2005(5)．
[41] 柏海涛，张建州．基于MPEG-4实时视频传输系统的设计[J]．网络纵横，2005(3)．
[42] 钟玉琢，王琪．MPEG-2运动图像压缩编码国际标准及MPEG的新进展[M]．北京：清华大学出版社，2002．
[43] 杨淑莹．VC++图像处理程序设计[M]，北京：清华大学出版社，北京交通大学出版社，2005．
[44] 樊星，顾伟康，叶秀清．多媒体会议中的快速实时自适应混音方案研究[J]．软件学报，2005(16)．
[45] 许伟．用定点DSP实现音频会议的数字混音方案[J]．数字信号处理器，2000(5)．
[46] Jun X., Ming T. S., Kousou K. Bit allocation for joint transcoding of multiple MPEG coded video streams[J]. IEEE Int Confon Multimedia and Expo. Tokyo, Japan: IEEE Society Press, 2001. 119-122.
[47] Sorial H., Lynch W. E., Vincent A. Joint bit allocation for MPEG encoding of multiple video sequences with minimum quality variation [J]. Proc IEEE Int Sym on Circuits and Systems. Geneva, Switzerland: IEEE Society P ress, 2000. 9-12.
[48] 曹理宇，姚丹霖．MPEG-4流媒体系统中的视音频同步[J]．计算机应用，2005(1)．
[49] 单红鹏．IP组播技术综述[J]．计算机世界日报，2001(5)．
[50] 陆其明．Windows Media编程导向[M]．北京：清华大学出版社，2005．
[51] Gregory J.Conklin; Gray S.GreenBaum. Video Coding For Streaming Media Delivery on the Internet[J]. IEEE TRANSCATION ON CIRCUITS AND SYSTEM FOR VIDEO THECHNOLOGY, 2001.
[52] KwangYong Kim; JinWoo Hong; ChuMin Park; HoeKyung Jung. The architecture of MPEG-4 based IPMP authoring system[J]. Multimedia and Expo, 2004. ICME'04. 2004 IEEE International Conference, 2004(6).
[53] MacKnight, R.B.; Timans, P.J.; Tay, S.-P.; Nenyei, Z. RTP application and technology options for the sub-45 nm nodes[J]. Advanced Thermal Processing of Semiconductors, 2004. RTP 2004. 12th IEEE International Conference, 2004(9).
[54] Mark Michaelis. COM+编程指南[M]．北京：机械工业出版社，2002．
[55] 蔡孟哲．COM／Active X实用技巧[M]．北京：中国铁道出版社，2001．
[56] 宋玲，陈燕．基于DirectShow实现网络音视频实时发送和回放[J]．电子技术应用，2006(1)．
[2] 孟祥清，王建民．软件视频会议的实现[J]．视听界广播电视技术，2006(4)．
[3] 詹承华．视频会议系统中多点显示单元的设计与实现[D]．陕西：西北工业大学软件工程系，2006．
[4] 涂友华．基于SIP的IP视频会议系统的研究[D]．北京：北京邮电大学计算机应用系，2006．
[5] 吴震．视频会议系统的研究与实现[D]．四川：电子科技大学软件工程系，2006．
[6] 史臻赜．一种新型的通用视频会议系统研究与设计[J]．计算机与信息技术，2005(10)．
[7] 张新盛，郑建彬．基于软交换的IP视频会议技术研究[J]．计算机与信息技术，2005(10)．
[8] 陆其明．DirectShow实务精选[M]．北京：北京科海电子出版社，2004．
[9] 陆其明．DirectShow开发指南[M]．北京：清华大学出版社，2005．
[10] 杨淑莹．图像模式识别—VC++技术实现[M]．北京：清华大学出版社，北京交通大学出版社，2005．
[11] 杨明川．层次化的宽带流媒体服务质量保证[J]．信息网络，2005(2)．
[12] 徐瑛．DirectShow技术在远程教学系统中的应用研究[J]．职业技术学院学报，2006(5)．
[13] 顾学飞．基于DirectShow的广域网P2P视频会话系统[J]．江南大学学报，2006(5)．
[14] 丰伟伟．DirectShow播放视频和MP3[J]．电脑编程技巧与维护，2006(12)．
[15] Cline, L.S.; Du, J.; Keany, B.; Lakshman, K.; Maciocco, C.; Putzolu, D.M. DirectShow~(TM) RTP support for adaptivity in networked multimedia applications[J]. Multimedia Computing and Systems, 1998. Proceedings. IEEE International Conference. 1998(7).
[16] Microsoft Corporation. DirectShow 9.0 Document[Z]. Microsoft, 2003.
[17] Microsoft Corporation. Microsoft MSDN library[Z], http://www.microsoft.com.
[18] 刘平，高胜法．一种基于DirectShow COM接口的多媒体流解码程序的设计[J]．山东科学，2002(5)．
[19] 王厚雪．基于目标识别的视音频合成传输系统设计[D]．天津：天津理工大学计算机系，2005．
[20] Viriyaphol, p.; Jittawiriyanukoon, C. Efficient routing with clue protocol for IP based networks[J]. Mobile Technology, Applications and Systems, 2005 2nd International Conference, 2005(10).
[21] Sengupta, S.; Kumar, V.; Saha, D. Switched optical backbone for cost-effective scalable core IP networks[J]. Communications Magazine, IEEE, 2003(6).
[22] Young-ho Lee; Ki-Won Kwon; Jin-Tea Kim; Chul-Dong Lee. IP development and management of IP DB enabling efficient system-on-chip design[J]. ASICs, 1999. AP-ASIC'99. The First IEEE Asia Pacific Conference, 1999(8).
[23] 鲁宏伟，龚雄斌，孔华锋．基于接收端应用的TCP流量控制策略的研究[J]．小型微型计算机系统，2006(12)．
[24] 苑春苗，杨清永，王闵．IP网上稳定可靠的视频流系统设计与实现[J]．计算机工程，2003(7)．
[25] 余萍．基于Internet的实时视频传输技术浅析[J]．电力系统通信，2002(4)．
[26] Kai Xu; Ye Tian; Ansari, N. TCP-Jersey for wireless IP communications[J]. Selected Areas in Communications, IEEE Journal, 2004(5).
[27] Ebrahimi-Taghizadeh, S.; Helmy, A.; Gupta, S. TCP vs. TCP: a systematic study of adverse impact of short-lived TCP flows on long-lived TCP flows[J]. INFOCOM 2005. 24th Annual Joint Conference of the IEEE Computer and Communications Societies. Proceedings IEEE, 2005(3).
[28] Girish Motwani; Gopinath, K. Evaluation of advanced TCP stacks in the iSCSI environment using simulation model[J]. Mass Storage Systems and Technologies, 2005. proceedings. 22nd IEEE/13th NASA Goddard Conference, 2005(3).
[29] 刘捷，龚根华．基于UDP的新型数据传输协议[J]．南昌大学学报(理科版)，2006(5)．
[30] 赵飞，叶震．UDP协议与TCP协议的对比分析与可靠性改进[J]．计算机技术与发展，2006(9)．
[31] Lam, P. P. -K.; Liew, S.C. UDP-Liter: an improved UDP protocol for real-time multimedia applications over wireless links[J]. Wireless Communication Systems, 2004. 1st International Symposium, 2004(9).
[32] Chang, R. K. C.; Fung, K. P. Transport layer proxy for stateful UDP packet filtering[J]. Computers and Communications, 2002. Proceedings. ISCC 2002. Seventh Internationsl Symposium, 2002(7).
[33] Mertz, F.; Engelke, U.; Vary, P.; Taddei, H.; Varga, I. Applicability of UDP-lite for voice over IP in UMTS networks[J]. Personal, Indoor and Mobile Radio Communications, 2005. PIMRC 2005. IEEE 16th Internatinal Symposium, 2005(9).
[34] 郑力明，张会汀．基于IP组播技术的分布式视频会议系统的设计与实现[J]．计算机工程与应用，2003(2)．
[35] 杨晓东，乐正友．IP组播技术及流式媒体在网上音视频广播中的应用[J]．现代电视技术，2002(4)．
[36] 王斌，刘增基．基于ACK和NACK的可靠组播传输协议的性能分析和比较[J]．电子学报，2001 (10)．
[37] Ki-ll Kim; Jeoung-Lak Ha; Eun-Hee Hyun; Sang-Ha Kim. Internet multicast provisioning issues for hierarchical architecture[J]. Networks, 2001. Proceedings. Ninth IEEE Internatinal Conference, 2001(10).
[38] Xue-Bin Liang. Matrix games in the multicast networks: maximum information flows with network switching[J]. Information Theory, IEEE Transactions, 2006(6).
[39] 王正勇，李永合．MPEG-4基于内容的视频编码研究[J]．四川大学学报(自然科学版)，2001(10)．
[40] 雷刚勇，解梅．基于DirectShow的MPEG-4远程视频监控系统[J]．视频技术应用与工程，2005(5)．
[41] 柏海涛，张建州．基于MPEG-4实时视频传输系统的设计[J]．网络纵横，2005(3)．
[42] 钟玉琢，王琪．MPEG-2运动图像压缩编码国际标准及MPEG的新进展[M]．北京：清华大学出版社，2002．
[43] 杨淑莹．VC++图像处理程序设计[M]，北京：清华大学出版社，北京交通大学出版社，2005．
[44] 樊星，顾伟康，叶秀清．多媒体会议中的快速实时自适应混音方案研究[J]．软件学报，2005(16)．
[45] 许伟．用定点DSP实现音频会议的数字混音方案[J]．数字信号处理器，2000(5)．
[46] Jun X., Ming T. S., Kousou K. Bit allocation for joint transcoding of multiple MPEG coded video streams[J]. IEEE Int Confon Multimedia and Expo. Tokyo, Japan: IEEE Society Press, 2001. 119-122.
[47] Sorial H., Lynch W. E., Vincent A. Joint bit allocation for MPEG encoding of multiple video sequences with minimum quality variation [J]. Proc IEEE Int Sym on Circuits and Systems. Geneva, Switzerland: IEEE Society P ress, 2000. 9-12.
[48] 曹理宇，姚丹霖．MPEG-4流媒体系统中的视音频同步[J]．计算机应用，2005(1)．
[49] 单红鹏．IP组播技术综述[J]．计算机世界日报，2001(5)．
[50] 陆其明．Windows Media编程导向[M]．北京：清华大学出版社，2005．
[51] Gregory J.Conklin; Gray S.GreenBaum. Video Coding For Streaming Media Delivery on the Internet[J]. IEEE TRANSCATION ON CIRCUITS AND SYSTEM FOR VIDEO THECHNOLOGY, 2001.
[52] KwangYong Kim; JinWoo Hong; ChuMin Park; HoeKyung Jung. The architecture of MPEG-4 based IPMP authoring system[J]. Multimedia and Expo, 2004. ICME'04. 2004 IEEE International Conference, 2004(6).
[53] MacKnight, R.B.; Timans, P.J.; Tay, S.-P.; Nenyei, Z. RTP application and technology options for the sub-45 nm nodes[J]. Advanced Thermal Processing of Semiconductors, 2004. RTP 2004. 12th IEEE International Conference, 2004(9).
[54] Mark Michaelis. COM+编程指南[M]．北京：机械工业出版社，2002．
[55] 蔡孟哲．COM／Active X实用技巧[M]．北京：中国铁道出版社，2001．
[56] 宋玲，陈燕．基于DirectShow实现网络音视频实时发送和回放[J]．电子技术应用，2006(1)．