视频监控中运动目标检测与异常分析技术的研究
|
摘要
近年来,随着中国经济的高速发展以及“平安城市”建设步伐的迈进,智能视频监控系统也开始在一些重要场合和公共领域迅速普及起来。它的异常分析功能可以自动判断目标行为,因此既能完成日常管理又能在异常情况发生的时候及时做出反应。本文对视频监控中的异常分析构建了一个完整的系统方案。首先用混合多高斯模型与基本背景减法相结合进行运动目标检测;然后用基于MeanShift算法的方法跟踪目标;最后在跟踪的过程中分析目标是否有禁区入侵异常、快速奔跑异常或遗留包裹异常。通过用多段不同实验视频进行算法仿真,本文验证了所用异常分析方法的有效性以及在实际应用中的可行性。
In recent years, with China’s rapid economic development as well as the pace of construction of“Green City”, intelligent video surveillance system has also begun to popularize in some important occasions and public domains.Its function of anomaly analysis can automatically determine the objective behavior,so it can not only complete the daily management but also respond timely when the anomalies occur.In this paper,a complete syetem solution about anomal analysis of video surveillance is built.First,detecting the moving object using the combining method of hybrid Multi-Gaussian Model and Basic Background Subtraction. Then,tracking the object using the method based on MeanShift algorithm.Finally,analyzing the object whether it has the anomalies of invasion,rapid running or leaving over the parcels.By simulating many different experimental videoes,the test verify the validity of the anomaly analysis methods and the feasibility in practical application.
In recent years, with China’s rapid economic development as well as the pace of construction of“Green City”, intelligent video surveillance system has also begun to popularize in some important occasions and public domains.Its function of anomaly analysis can automatically determine the objective behavior,so it can not only complete the daily management but also respond timely when the anomalies occur.In this paper,a complete syetem solution about anomal analysis of video surveillance is built.First,detecting the moving object using the combining method of hybrid Multi-Gaussian Model and Basic Background Subtraction. Then,tracking the object using the method based on MeanShift algorithm.Finally,analyzing the object whether it has the anomalies of invasion,rapid running or leaving over the parcels.By simulating many different experimental videoes,the test verify the validity of the anomaly analysis methods and the feasibility in practical application.
引文
[1]廖国华.视频监控中的关键技术.通信与信息技术,2008,4:58～61
[2]孔晓东.智能视频监控技术研究:[博士学位论文].上海:上海交通大学,2008
[3]王德彪,程大章.基于视频图像的运动目标检测技术.消防与安防,2008,22:25～27
[4]Rseneau S,Cooperstock J.Real-time Image Segmentation for Action Recognition.Proc IEEE Pacific Rim Conference on Communications, Computers and SignalProcessing,1999:86～89
[5]Stauffer C,Grimson W.Learning patterns of activity using real time tracking. IEEE Transactions on Pattern Analysis and MachineIntelligence,2000, 22(8):747～767
[6]Fernando D T,Michael J B.A framework for robust subspace learning. International Journal of Computer Vision,2003,54(7):117～142
[7]Papageorgiou C,Oren M,Poggio T.A general framework for object detection. IEEE International Conference on Computer Vision,1998:555～562
[8]Rowley H,Baluja S,Kanade T.Neural network-based face detection. IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(1):23～38
[9]Barron J L,Fleet D J,Beauchemin S S.Systems and experiment performance of optical flow techniques.International Journal of Computer Vision,1994,12(1):43～77
[10]陈建林,丁永生,郝矿荣,张淑平.基于目标跟踪的并联机器人视觉位姿检测.计算机工程,2009,35(18):200～203
[11]郭孔辉,吕济明,丁海涛,郭文鑫.基于MATLAB的车辆组件模型库的设计与实现.吉林大学学报,2006,36(6):866～870
[12]代科学,李国辉,武德峰.聚类挖掘在监控视频中的应用.测控技术,2006,25(10):16～19
[13]章毓晋.图像分割.北京:科学出版社,2001,35～45
[14]Toyama K,Krumm J,Burtnitt B,Meyers B.Principles and Practie of background maintenance.Proeeedings IEEE International Conference Computer Vision,1999:255～261
[15]赵文哲,秦世引.视频运动目标检测方法的对比分析.科技导报,2009,27(10):64～70
[16]Jain R,Nagel H H.On the analysis of accumulative difference picture from image sequences of real world scenes.IEEE Trans.on PAMI,1979,1(2):206～214
[17]夏永泉,李卫丽,甘勇,张素智.智能视频监控中的运动目标检测技术研究.通信技术,2009,6(42):185～187
[18]邱丹丹,马洪超,杨耘.基于多源卫星图像融合的水坝检测方法的研究.遥感学报,2006,10(4):449～455
[19]邓湘金,彭海良.一种基于遥感图像的机场检测方法.测试技术学报,2002,16(2): 96～99
[20]Collins R,LIPton A,Fujiyoshi H,Kanade T.Algorithms for Cooperative Multisensor Surveillance,proc.IEEE,2001,89(10):1456～1477
[21]ZivkovieZ.lmproved adaptive Gaussian mixture model for background Subtraction,Proceedings of the 17th International Collference on Pattern Recognition,2004,2:28～31
[22]Chris Stauffer,W.Eric,L.Grimson.Learning Patterns of Activity Using Real-time Tracking.IEEE Trans.On Patten Analysis and Machine Intelligence,2000,22(8):747～757
[23]Hall D,Nascimento J,Ribeiro P.Comparison of target detection algorithms using adaptive background models:2005,46:1～8
[24]岳佳,王士同.高斯混合模型聚类中EM算法及其初始化问题的研究.微计算机信息,2006,22(11):244～246
[25]Barron J,Fleet D,Beauchemin S.Performance of optical flow techniques. International Journal of computer Vision, 1994,12(1):42～77
[26]沈海浪,平西建,许志勇.一种基于背景重建的运动目标检测算法.信息工程大学学报,2004,5(3):69～71
[27]邵文坤,黄爱民,韦庆.目标跟踪方法综述.影像技术,2006,1:17～20
[28]艾金慰,刘克.视频序列中运动目标跟踪新方法.北京科技大学学报,2006,28(2):195～198
[29]董春利,董育宁,王莉.活动轮廓模型目标跟踪算法综述.计算机工程与应用,2008,44(34):208～212
[30]张昊,战华,道银.使用粒子滤波和差分进化法实现轮廓跟踪.光电工程,2006,33(2):29～32
[31]张崇友,董慧颖,兰利宝.基于粒子滤波的相关跟踪算法研究.沈阳理工大学学报,2008,27(1):6～9
[32]潘薇,游志胜,吴鹍等.基于模糊聚类和卡尔曼滤波的运动目标检测[J].计算机应用,2005,25(1):123～124
[33]魏娟丽,翟社平,王万诚.视频序列中人体运动目标的检测与跟踪研究.计算机应用与软件,2006,23(4):139～141
[34]任伟建,山茂泉,谢锋,王文东.基于粒子滤波和贝叶斯估计的目标跟踪.大庆石油学院学报,2008,32(3):67～70
[35]张宏志,张金换,岳卉,黄世霖.基于CamShift的目标跟踪算法.计算机工程与设计,2006,27(11):2012～2014
[36]万力,梁昌斌.智能视频监控系统中一种基于MeanShift的目标跟踪计数法.智能建筑,2008,94:59～61
[37]雷玉堂.金融安防监控必须向智能化发展.第十一届安博会权威论坛,2007:126～131
[38]谭墍元,吴成东,周芸,侯俊,王俏俏.智能图像监控系统异常目标检测算法研究.机电工程,2009,26(3):12～15
[39]宋磊,黄祥林,沈兰荪.视频监控中的快速异常检测与分析.高技术通讯,2004,1:1～5
[40]陈宜稳,王威,王润生.基于视频区域特征的行人异常行为检测.计算机应用,2007,27(10):2610～2611
[41]Beynon M,Hook D,Seibert M.Detecting abandonded packages in a Multi-camera video surveillance system[C].Proceedings of the IEEE Conference on Advanced Video and Signal Based Surveillance,2003:221～228
[42]张春杰,李继云.基于HMM的无主包裹识别.计算机技术与发展,2009,19(3):28～31
[43]胡正平,吴燕,张晔.基于几何分析的支持向量机快速训练与分类算法.中国图象图形学报,2007,1(12):82～86
[44]潘泉,程咏梅,杜亚娟,张洪才.离散不变矩算法及其在目标识别中的应用.电子与信息学报,2001,23(1):30～36
[45]陈天华.面向对象程序设计与Visual C++6.0教程.北京:清华大学出版社,2006,339～390
[2]孔晓东.智能视频监控技术研究:[博士学位论文].上海:上海交通大学,2008
[3]王德彪,程大章.基于视频图像的运动目标检测技术.消防与安防,2008,22:25～27
[4]Rseneau S,Cooperstock J.Real-time Image Segmentation for Action Recognition.Proc IEEE Pacific Rim Conference on Communications, Computers and SignalProcessing,1999:86～89
[5]Stauffer C,Grimson W.Learning patterns of activity using real time tracking. IEEE Transactions on Pattern Analysis and MachineIntelligence,2000, 22(8):747～767
[6]Fernando D T,Michael J B.A framework for robust subspace learning. International Journal of Computer Vision,2003,54(7):117～142
[7]Papageorgiou C,Oren M,Poggio T.A general framework for object detection. IEEE International Conference on Computer Vision,1998:555～562
[8]Rowley H,Baluja S,Kanade T.Neural network-based face detection. IEEE Transactions on Pattern Analysis and Machine Intelligence,1998,20(1):23～38
[9]Barron J L,Fleet D J,Beauchemin S S.Systems and experiment performance of optical flow techniques.International Journal of Computer Vision,1994,12(1):43～77
[10]陈建林,丁永生,郝矿荣,张淑平.基于目标跟踪的并联机器人视觉位姿检测.计算机工程,2009,35(18):200～203
[11]郭孔辉,吕济明,丁海涛,郭文鑫.基于MATLAB的车辆组件模型库的设计与实现.吉林大学学报,2006,36(6):866～870
[12]代科学,李国辉,武德峰.聚类挖掘在监控视频中的应用.测控技术,2006,25(10):16～19
[13]章毓晋.图像分割.北京:科学出版社,2001,35～45
[14]Toyama K,Krumm J,Burtnitt B,Meyers B.Principles and Practie of background maintenance.Proeeedings IEEE International Conference Computer Vision,1999:255～261
[15]赵文哲,秦世引.视频运动目标检测方法的对比分析.科技导报,2009,27(10):64～70
[16]Jain R,Nagel H H.On the analysis of accumulative difference picture from image sequences of real world scenes.IEEE Trans.on PAMI,1979,1(2):206～214
[17]夏永泉,李卫丽,甘勇,张素智.智能视频监控中的运动目标检测技术研究.通信技术,2009,6(42):185～187
[18]邱丹丹,马洪超,杨耘.基于多源卫星图像融合的水坝检测方法的研究.遥感学报,2006,10(4):449～455
[19]邓湘金,彭海良.一种基于遥感图像的机场检测方法.测试技术学报,2002,16(2): 96～99
[20]Collins R,LIPton A,Fujiyoshi H,Kanade T.Algorithms for Cooperative Multisensor Surveillance,proc.IEEE,2001,89(10):1456～1477
[21]ZivkovieZ.lmproved adaptive Gaussian mixture model for background Subtraction,Proceedings of the 17th International Collference on Pattern Recognition,2004,2:28～31
[22]Chris Stauffer,W.Eric,L.Grimson.Learning Patterns of Activity Using Real-time Tracking.IEEE Trans.On Patten Analysis and Machine Intelligence,2000,22(8):747～757
[23]Hall D,Nascimento J,Ribeiro P.Comparison of target detection algorithms using adaptive background models:2005,46:1～8
[24]岳佳,王士同.高斯混合模型聚类中EM算法及其初始化问题的研究.微计算机信息,2006,22(11):244～246
[25]Barron J,Fleet D,Beauchemin S.Performance of optical flow techniques. International Journal of computer Vision, 1994,12(1):42～77
[26]沈海浪,平西建,许志勇.一种基于背景重建的运动目标检测算法.信息工程大学学报,2004,5(3):69～71
[27]邵文坤,黄爱民,韦庆.目标跟踪方法综述.影像技术,2006,1:17～20
[28]艾金慰,刘克.视频序列中运动目标跟踪新方法.北京科技大学学报,2006,28(2):195～198
[29]董春利,董育宁,王莉.活动轮廓模型目标跟踪算法综述.计算机工程与应用,2008,44(34):208～212
[30]张昊,战华,道银.使用粒子滤波和差分进化法实现轮廓跟踪.光电工程,2006,33(2):29～32
[31]张崇友,董慧颖,兰利宝.基于粒子滤波的相关跟踪算法研究.沈阳理工大学学报,2008,27(1):6～9
[32]潘薇,游志胜,吴鹍等.基于模糊聚类和卡尔曼滤波的运动目标检测[J].计算机应用,2005,25(1):123～124
[33]魏娟丽,翟社平,王万诚.视频序列中人体运动目标的检测与跟踪研究.计算机应用与软件,2006,23(4):139～141
[34]任伟建,山茂泉,谢锋,王文东.基于粒子滤波和贝叶斯估计的目标跟踪.大庆石油学院学报,2008,32(3):67～70
[35]张宏志,张金换,岳卉,黄世霖.基于CamShift的目标跟踪算法.计算机工程与设计,2006,27(11):2012～2014
[36]万力,梁昌斌.智能视频监控系统中一种基于MeanShift的目标跟踪计数法.智能建筑,2008,94:59～61
[37]雷玉堂.金融安防监控必须向智能化发展.第十一届安博会权威论坛,2007:126～131
[38]谭墍元,吴成东,周芸,侯俊,王俏俏.智能图像监控系统异常目标检测算法研究.机电工程,2009,26(3):12～15
[39]宋磊,黄祥林,沈兰荪.视频监控中的快速异常检测与分析.高技术通讯,2004,1:1～5
[40]陈宜稳,王威,王润生.基于视频区域特征的行人异常行为检测.计算机应用,2007,27(10):2610～2611
[41]Beynon M,Hook D,Seibert M.Detecting abandonded packages in a Multi-camera video surveillance system[C].Proceedings of the IEEE Conference on Advanced Video and Signal Based Surveillance,2003:221～228
[42]张春杰,李继云.基于HMM的无主包裹识别.计算机技术与发展,2009,19(3):28～31
[43]胡正平,吴燕,张晔.基于几何分析的支持向量机快速训练与分类算法.中国图象图形学报,2007,1(12):82～86
[44]潘泉,程咏梅,杜亚娟,张洪才.离散不变矩算法及其在目标识别中的应用.电子与信息学报,2001,23(1):30～36
[45]陈天华.面向对象程序设计与Visual C++6.0教程.北京:清华大学出版社,2006,339～390