摘要
视频监控系统以其直观、方便、信息内容丰富而成为现代安防系统发展的主流。其中,数字视频监控系统采用了多媒体技术、网络传输技术及数字视频图像处理技术等众多新技术,符合当前信息社会数字化、网络化和智能化的发展趋势,被广泛应用于各行各业。
对视频图像进行运动目标检测是实现数字视频监控系统智能化的基础。本文在详细分析、研究了目前常用的运动目标检测方法和技术实现后,针对静止背景下的单摄像头监控,采用相邻帧差法对视频监控图像进行运动检测,在保证记录到有事件发生的关键场景的同时,有效提高了存储空间的利用率及事后检索的方便性,从而降低了监控人员的工作强度。
本文首先介绍了数字视频监控系统中所采用的关键技术,然后详细分析了系统所要实现的目标,从而确定了系统实现要采用的技术路线。最终设计并实现了一个C/S通信模式下的数字视频监控系统,由视频采集、运动检测、视频处理、网络传输及视频的接收、播放模块组成。系统中对于要求高可靠性的控制指令,选用TCP协议;而对实时性要求高的视频流数据的通信则选用基于UDP协议的IP组播实现。系统在现场端利用DirectShow技术实现对所捕获视频的播放,视频数据经运动检测、MPEG-4压缩处理后,利用位于UDP协议之上的RTP/RTCP协议将编码后的视频流发送到远程接收端。接收端负责接收和解码播放发送来的视频数据,并在接收过程中采用双缓冲技术来缓解网络延迟和抖动带来的影响。视频数据发送端和接收端均采用多线程技术实现。
论文所述的系统,采用DirectShow技术,实现对所捕获视频的播放,因此使得系统具有与底层硬件无关的通用性,而运动目标检测技术的加入,为系统功能柔性扩充和完善奠定了极为重要的基础。
As a growing tendency of modern defence system, Video Surveillance System has charactors of visibility, convenience and abundant in content. Many advantaged technologies are applicated in digital video surveillance system including multimedia, network transmission and digital video imagery processing technology etc. That made the system according with the digital, network and intelligent developing trend of current information society. So it can be applied to various fields.
To detect the moving objects from raw video is the base of implementing intelligent surveillance system. A frame differencing algorithm is accomplished to detect moving objects under the stationary background with single camera after researching in detail three conventional approaches of moving object detection in this thesis. Through using this approach, not only the sensitive scene can be recorded, but also the storage and the labor of monitoring can be saved simultaneously.
This thesis presents the current key technologies of Digital Video Surveillance System firstly, and then introduces the technical routine that has been choiced. Finally, the system is designed and built, which be consisted of the client and server side including the module of video data capture, moving object detection, video data compressor, net transmission and video data receiver, decoder and playback. TCP is used for the control command that need the high reliability, and IP Multicast based on UDP for the transmission of video data. In the server, the captured video data is displayed on the screen, and the moving object is detected simultaneously, the data is encoded by MPEG-4 compression standard, then the compressed data will be multicasted to the client based on RTP. The client is responsible for receiving, decoding and playing the video imagery. Bi-buffer technology is achieved in receiving the RTP packet to alleviate the delay of network. The technology of Multithread is used in both sender side and receiver side.
The digital video surveillance system which is described in the thesis is possessed of hardware-independent generality for the using of DirectShow Framework. Further, the application of moving object detection lays the foundation for the expandability and improvement of the system function.
引文
[1]陈岩峰.网络视频监控的灵魂:视频管理软件.智能建筑与城市信息[J].2007年第3期.18-19.
[2]中国网.平安城市建设带动视频监控市场迅速增长[OL].http://www.china.com.cn/tecb./txt/2007-05/01/content_8201229.htm,2007.5.1.
[3]中安网.2007年平安城市建设如火如荼[OL].http://www.cps.com.cn/news/Html/newcenter/TBBD/1847220070619.htm,2007.6.19.
[4]李波.数字视频监控系统:多一双眼睛看世界[OL].http://tech.sina.com.cn/oi/2001-10-30/89889.shtml,2001.10.30
[5]Danielson P.Video surveillance for the rest of us:proliferation,privacy,and ethics education[J].Technology and Society,2002.(ISTAS'02).2002 International Symposium on 6-8 June 2002:162-167.
[6]R Collins,A Lipton et al.A system for video surveillance and monitoring:VSAM final report.CMU-RI-TR-00-12,Robotic Institute Carnegie Mellon University,2000.
[7]R Collins,A Lipton,and T.Kanade.Introduction to the special section on video surveillance[C].IEEE Trans.Pattern Analysis and Machine Intelligence,22(8):745-746,August 2000.
[8]深圳市永研科技发展有限公司.NEC SmartCatch旧金山机场监控解决方案[OL].http://info.secu.hc360.com/2007/06/271957117082.shtml,2007.6.27.
[9]Frost,Sullivan.全球视频监控市场的现状及未来发展[J].A&S:国际中文版,2006.9:96-99.
[10]北京兆维泰奇科技有限公司.智能视频安防设备正在进入城市监控市场[OL].http://info.secu.hc360.com/2007/06/232202116822.shtml,2007.6.23.
[11]中安网.思科、索尼、同天公司共同推出全IP网络视频监控解决方案[OL].http://www.cps.com.cn/news/Html/newcenter/QYDT/0547520040602.htm,2004.6.2.
[12]四维科技 刘玮玮.Visual C++视频/音频开发实用工程案例精选[M],北京:人民邮电出版社,2004.65-66.
[13]辛长安,梅林.VC++编程技术与难点剖析[M].北京:清华大学出版社,2002.4.
[14]陈兵旗,孙明.Visual C++实用图像处理专业教程[M].北京:清华大学出版社,2004.3.
[15](美)Jon Bates,Tim Tompkins著,何键辉,董方鹏译.实用Visual C++6.0教程[M].北京:清华大学出版社,2000.8.
[16]王文庆,毛文林,张利飞.基于MPEG-4的视频监控系统模型及实现[J].计算机工程.2006,32(11):237-239.
[17]高焕兵.C/S模式视频监控系统的设计与实现[J].成都信息工程学院学报,2003.12.386-389.
[18]Tekalp A M.Digital Video Processing.America:Prentice Hall.Inc,1996.
[19]SHENG Ting-zhi,FANG Zhi-wen.Digital Image Processing and Pattern Recognition.Beijing:Beijing University of Technology and Science Press,1998(Ch).
[20]LI Zun-ming.Basic Theory of TV Image Auto Track.Beijing:Naional Defence Industry Press,1998(Ch).
[21]胡俊,苏祥芳,刘立海等.图像序列运动检测算法的研究及其应用[J].武汉:武汉大学学报(自然科学版).2000.10.第46卷,第5期.613-616.
[22]A Neri,S Colonnese,G Russo et al.Automatic moving object and background separation[J],Signal Processing,1998,66:219-232.
[23]R Mech,M Wollborn.A noise robust method for segmentation of moving objects in video sequences[C],In:Proc International Conference on Acoustics,Speech and Signal Processing,1997,41-45.
[24]Ramesh Jain,H H Nagel.On the Analysis of Accumulative Difference Picture from Image Sequences of Real World Scenes.IEEE Trans.on PAMI.1979,1(2).206-214.
[25]Lipton A.J,Fujiyoshi,H,Patil,R.S.Moving target classification and tracking from real-time video[C],Proc.IEEE Workshop on Applications of Computer vision,1998 8.14.
[26]张广军.机器视觉(第二版)[M].北京:科学出版社,2005.
[27]朱颖,江泽涛.基于累积绝对差的图像与交叉墒分割的运动目标检测与定位[J].计算机与现代化.2005,8(8).75-77.
[28]F Moscheni,S Bhattacharjee,M Kunt.Spatial temporal segmentation based on region merging[J],IEEE Transaction on pattern analysis and machine intelligence,1998,20:897-915.
[29]Haritaoglu I,Harwood D,Davis L.W4:Real-time surveillance of people and their activities[C].IEEE Trans Pattern Analysis and Machine Intelligence,2000,22(8):809-830.
[30]Stauffer C,Grimson W.E.L.Adaptive background mixture models for real-time tracking[C],In Computer Vision and Pattern Recognition,246-252.
[31]Anurag Mittal,Nikons Paragios.Motion-based background subtraction using adaptive kernel density estimation In CVPR'04,volume 2:302-309.
[32]Horprasert T,Harwood D,Davis L.A Statistical approach for real-time robust background subtraction and shadow detection[C],In:Proc.ICCV'99.1-19.
[33]Arseneau S,Cooperstock J.Real-time image segmentation for action recognition[C].In:Proc IEEE Pacific Rim Conference on Communications,Computers and Signal Processing,Victoria,Canada,1999.86-89.
[34]Barron J,Fleet D,Beauchemin S.Performance of optical flow techniques[J],International journal of computer vision,1994:12(1):42-77.
[35]Mae,Y.,et al.Optical Flow Based Realtime object tracking by active vision system[C],In:Proe.2 nd Japan-France Congress on Mechatronics,1994,545-548.
[36]王向阳.多媒体信息处理技术原理[M].北京:科学出版社,2005.
[37]张冬明.复杂度可分级的视频编码技术研究[博士学位论文],中国科学院计算技术研究所,2006.
[38]高鹏.视频会议系统与MPEG-4标准的研究[博士后出站报告],中国科学院计算技术研究所计算机科学与技术,1999.
[39]余兆明,李晓飞,陈来春编著.MPEG标准及其应用[M].北京:北京邮电大学出版社,2002.11.
[40]沈兰荪编著.视频编码与低速率传输[M].北京:电子工业出版社,2001.12.
[4l]吕行.基于DirectShow的智能视频监控系统研究[硕士学位论文].西安:西北工业大学,2007.
[42]DivX和XviD不能不说的故事[OL].http://virus.blog.51cto.com/51437/22285,2007.4.1
[43]DivX/XviD[OL].http://baike.baidu.com/view/471666.htm,2008.1.30.
[44]鲁文士.多媒体网络与应用[M].北京:清华大学出版社,2002.
[45](美)特南鲍姆(Tanenbaum,A.S.)著;潘爱民译.计算机网络:第四版(Computer Networks,Fourth Edition)[M].北京:清华大学出版社,2004.8.
[46]张文俊等主编.数字媒体技术基础[M].上海:上海大学出版社,2007.3.70-79.
[47]朱鹏,李春文.基于RTP的网络视频传输系统的设计与实现[J].计算机工程与应用,2003,26:18-20.
[48]詹青龙主编.网络视频技术及应用=Technology of Web-Video and Application[M].西安:西安电子科技大学出版社,2004.1.11-12.
[49]S.Casner,R.Frederick,V.Jacobson,RFC3550,RTP:A Transport Protocol for Real-Time Applications,2003.
[50]H.Schulzrinne,S.Casner,RFC3551,RTP Profile for Audio and Video Conferences with Minimal Control,2003.
[51]H.Schulzrinne,AT&T Bell Laboratories(Murray Hill).RTP.MCNC 2nd Packet Video Workshop.December 1992.
[52]Linda S.Cline,John Du,Bernie Keany,et al.DirectShow RTP support for adaptivity in networked multimedia applications,IEEE International Conferenccon Multimedia Computing and Systems,1998:1-15.
[53]牛少平.计算机视频监控技术研究[硕士学位论文],西安:西北工业大学,2006.
[54]马川,刘志镜.视频实时传输的RTP实现.现代电子技术[J].2004(20).41-43.
[55]曹加恒,李晶.新一代多媒体技术与应用[M].武汉:武汉大学出版社,2006.1.
[56]阮新新主编.多媒体技术与应用[M].北京:科学出版社,2006.
[57]陆其明,DirectShow实务精选[M].北京:科学出版社,2004.
[58]汪晓平,刘韬等.Visual C++6.0开发网络典型应用实例导航[M].北京:人民邮电出版社,2005.9.
[59]张越等.Visual C++网络程序设计实例详解[M].北京:人民邮电出版社,2006.7.
[60]李树前.基于MPEG_4的网络视频监控技术研究与实现[硕士学位论文],南京:南京航空航天大学,2007.
[61]陆其明,DirectShow开发指南[M].北京:清华大学出版社,2003.
[62]Qiming Lu.DirectShow- a technical introduction[OL].http://www.programsalon.com/downloads74/ebook/detai1264507.html,2007.4.6.
[63]周彪.网络视频监控系统的研究和开发[硕士学位论文],吉林:吉林大学,2007.
[64]汪德洋.智能视频监控关键技术研究[硕士学位论文].西安:西安电子科技大学,2005.
[65]赵新荣.网络视频传输及动目标检测技术的研究[硕士学位论文].南京:南京航空航天大学,2007.
[66]DivX编码软件[OL].http://file.mydrivers.com/tools/others/DivX502Bundle.exe,2007.8.
[67]Jori Liesenborgs.JRTPLIB 3.5.2[OL].http://research.edm.luc.ac.be/jori/jrtplib/jrtplib.pdf,2006.3.26.
[68](美)斯托林斯(Stallings,W.)著,林琪,阎慧译.计算机网络与Internet协议和技术[M].北京:中国电力出版社.2005.