摘要
随着计算机视觉的发展,智能视频监控得到了更广泛的发展。特别是随着人们对公共安全意识的提高,人们对人群场景中的异常行为检测越来越关注,对于异常检测的研究也越来越多。本文就智能视频监控中的人群异常检测展开工作,分析了几种常见人群异常行为,并相对的提出了一些异常检测方法。因为本文研究中高密度场景中的异常行为,因此首先要对场景中人群的密度进行分类,再对高密度的场景检测是否有异常事件的发生。
对于人群密度估计问题,本文采用最小二乘法线性拟合和基于图像纹理分析的两种方法。线性拟合方法,利用前景检测方法,提取前景目标的像素数,人工统计场景中人群数目,利用最小二乘法线性拟合,此方法计算复杂度低,但该方法在人群遮挡严重时效果不理想。基于图像纹理的人群密度估计方法依靠不同密度的人群所体现的纹理信息不同,首先提取人群前景目标,然后计算前景目标的灰度共生矩阵,运用支持向量机方法进行密度的分类。该方法能解决人群出现遮挡的问题,而且由于只计算了前景目标的纹理信息,排除了背景的影响,因此分类结果的正确率得到了保障。
针对传统基于跟踪的异常检测方法无法适用的拥挤人群场景,提出一种根据单元格速度和前景像素数(大小)及其运动方向是否具有刚性运动特性来判别异常的检测方法。为了只分析前景目标忽略不相关的背景,首先将输入帧进行前景分割,再将输入帧分割成不重叠的单元格,通过计算单元格中前景像素的光流提取每个单元格的运动特征来判定异常的发生。其中,速度特征可以检测出速度过快的异常情况。为了区分出车和因人群走近而形成的大目标,提出运动方向统计的方法。实验表明此方法在较短的时间内具有较好的检测效果。针对群殴和人群聚集事件,利用人群密度、运动强度、运动方向方差和人群面积变化等特征进行检测,对于人群聚集事件后的后续事件也有一定的预测效果。
With the development of computer vision, intelligent video surveillance has been developed more widely. Especially with the improvement of people's consciousness of public security, people have paid more and more attention to the abnormal behavior detection research in crowed scenes. In this thesis, we focus on the abnormal behavior detection of crowed scene in intelligent video surveillance system, analyze several common crowed abnormal behaviors, and put forward some relative anomaly detection method. As we study the abnormal behavior of high density scene, we should classify the scene crowed density firstly, then detect the abnormal events in the high density crowed scene.
For the crowd density estimation problem, this thesis uses the method of least squares linear fit and the method based on image texture analysis. Linear fitting method, use foreground detection method to extract foreground pixels number of targets, count the number of population in the crowed scene artificially, use the method of least squares linear fit, this method has low computational complexity, but the method can't be used in the crowd scenes where collision happens. Crowd density estimation method based on image texture information reflects the different density crowd, Firstly it extracts the crowd foreground object, and then calculates the foreground object of the gray level co-occurrence matrix, using support vector machine to classify density. This method can be used in the crowd scenes where collision happens. As this method just calculates foreground objects texture information, excludes the influence of the background, the correct rate of classification results is reliable.
An efficient anomaly detection technique is proposed based on the cell speed as well as the number of foreground pixels and the movement direction. It is capable of dealing with crowded scenes where the traditional tracking based approaches tend to fail. Initial foreground segmentation of the input frames confines the analysis to foreground objects and effectively ignores irrelevant background dynamics. Input frames are split into non-overlapping cells, followed by extracting motion features by computing the optical flow of the foreground pixels to discriminate the normal and the Anomaly. Speed characteristics can detect high-speed abnormalities. In order to distinguish between vehicle and large object formed by people walking close to each other, this thesis puts forward the statistical method of the direction of movement. Experiments demonstrate that the proposed method has better detecting effect in a short period of time. The fighting of a large number of people and people gathering event can be detected by characteristics such as crowed density, motion intensity, direction of motion and crowed area change. We also can predict the things after people gathering event happens.
引文
[1]Khoudour L, Leclercq T, Bruyelle J L, et al. Local camera network for surveillance in Newcastle Airport[J].Intelligence Distributed Surveillance Systems, IEE Symposium on.2003.
[2]秦莉娟.基于内容的自动视频监控研究[D].浙江：浙江大学,2006.
[3]钟志,徐扬生,石为人,等.群体异常检测[J].仪器仪表学报,2007,28(4)：614-620.
[4]Mehran R, Oyama A, Shah M. Abnormal crowd behavior detection using social force model[C]//Computer Vision and Pattern Recognition,2009. CVPR 2009. IEEE Conference on. IEEE,2009:935-942.
[5]Adam A, Rivlin E, Shimshoni I, et al. Robust real-time unusual event detection using multiple fixed-location monitors[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on,2008,30(3):555-560.
[6]Zhan B, Monekosso D N, Remagnino P, et al. Crowd analysis:a survey[J]. Machine Vision and Applications,2008,19(5-6):345-357.
[7]Haering N, Venetianer P L, Lipton A. The evolution of video surveillance:an overview[J]. Machine Vision and Applications,2008,19(5-6):279-290.
[8]Tu P, Sebastian T, Doretto G, et al. Unified crowd segmentation[M]//Computer Vision-ECCV 2008. Springer Berlin Heidelberg,2008:691-704.
[9]Marques J S, Jorge P M, Abrantes A J, et al. Tracking groups of pedestrians in video sequences[C]//Computer Vision and Pattern Recognition Workshop,2003. CVPRW03. Conference on. IEEE,2003,9:101-101.
[10]Kratz L, Nishino K. Anomaly detection in extremely crowded scenes using spatio-temporal motion pattern models[C]//Computer Vision and Pattern Recognition,2009. CVPR 2009. IEEE Conference on. IEEE,2009:1446-1453.
[11]Andrade E L, Blunsden S, Fisher R B. Modelling crowd scenes for event detection[C]//Pattern Recognition,2006. ICPR 2006.18th International Conference on. IEEE,2006,1:175-178.
[12]Black M J. Explaining optical flow events with parameterized spatio-temporal models[C]//Computer Vision and Pattern Recognition,1999. IEEE Computer Society Conference on. IEEE,1999,1.
[13]Haritaoglu I, Harwood D, Davis L S. W4:real-time surveillance of people and their activities[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on,2000,22(8):809-830.
[14]T. Olson and F. Brill, "Moving object detection and event recognition algorithms for smart cameras," Proc. DARPA Image Understanding Workshop, pp.159-175, 1997.
[15]Zhao T, Nevatia R. Tracking multiple humans in complex situations[J]. Pattern Analysis and Machine Intelligence, IEEE Transactions on,2004,26(9): 1208-1221.
[16]Benezeth Y, Jodoin P M, Saligrama V, et al. Abnormal events detection based on spatio-temporal co-occurences[C]//Computer Vision and Pattern Recognition, 2009. CVPR 2009. IEEE Conference on. IEEE,2009:2458-2465.
[17]Kim J, Grauman K. Observe locally, infer globally:a space-time MRF for detecting abnormal activities with incremental updates[C]//Computer Vision and Pattern Recognition,2009. CVPR 2009. IEEE Conference on. IEEE,2009: 2921-2928.
[18]胡波,李晓华,沈兰荪.场景监控中的人群密度估计[J].电路与系统学报,2006,12(6)：19-22
[19]Yin J H, Velastin S A, Davies A C. Image processing techniques for crowd density estimation using a reference image[M]//Recent Developments in Computer Vision. Springer Berlin Heidelberg,1996:489-498.
[20]Wu X, Ou Y, Qian H, et al. A detection system for human abnormal behavior[C]//Intelligent Robots and Systems,2005.(IROS 2005).2005 IEEE/RSJ International Conference on. IEEE,2005:1204-1208.
[21]Wang S, Miao Z. Anomaly detection in crowd scene[C]//Signal Processing (ICSP),2010 IEEE 10th International Conference on. IEEE,2010:1220-1223.
[22]林佳乙,于哲舟,张健,等.基于背景差分法和帧间差分法的视频运动检测[J].2008中国仪器仪表与测控技术进展大会论文集(Ⅲ),2008.
[23]Naito T, Ito T, Kaneda Y. The obstacle detection method using optical flow estimation at the edge image[C]//Intelligent Vehicles Symposium,2007 IEEE. IEEE,2007:817-822.
[24]Hu W, Tan T, Wang L, et al. A survey on visual surveillance of object motion and behaviors[J]. Systems, Man, and Cybernetics, Part C:Applications and Reviews, IEEE Transactions on,2004,34(3):334-352.
[25]Ko T. A survey on behavior analysis in video surveillance for homeland security applications[C]//Applied Imagery Pattern Recognition Workshop,2008. AIPR'08. 37th IEEE. IEEE,2008:1-8.
[26]龙辉平,习胜丰,侯新华.实验数据的最小二乘拟合算法与分析[J].计算技术与自动化,2008,27(3)：20-23.
[27]Davies A C, Yin J H, Velastin S A. Crowd monitoring using image processingfJ]. Electronics& Communication Engineering Journal,1995,7(1):37-47.
[28]Cho S Y, Chow T W S, Leung C T. A neural-based crowd estimation by hybrid global learning algorithm[J]. Systems, Man, and Cybernetics, Part B:Cybernetics, IEEE Transactions on,1999,29(4):535-541.
[29]Chow T W S, Cho S Y. Industrial neural vision system for underground railway station platform surveillance [J]. Advanced Engineering Informatics,2002,16(1): 73-83.
[30]Marana A N, Velastin S A, Costa L F, et al. Estimation of crowd density using image processing[C]//Image Processing for Security Applications (Digest No.: 1997/074), IEE Colloquium on. IET,1997:11/1-11/8.
[31]张学工.关于统计学习理论与支持向量机[J].自动化学报,2000,26(1)：32-42.
[32]Wu X, Gong H, Chen P, et al. Surveillance robot utilizing video and audio information[J]. Journal of Intelligent and Robotic Systems,2009,55(4-5): 403-421.
[33]Zhang J, Liu Z. Detecting abnormal motion of pedestrian in video[C] //Information and Automation,2008. ICIA 2008. International Conference on. IEEE,2008:81-85.
[34]Yin J, Yang Q, Pan J J. Sensor-based abnormal human-activity detection[J]. Knowledge and Data Engineering, IEEE Transactions on,2008,20(8): 1082-1090.
[35]Foroughi H, Rezvanian A, Paziraee A. Robust fall detection using human shape and multi-class support vector machine[C]//Computer Vision, Graphics& Image Processing,2008. ICVGIP'08. Sixth Indian Conference on. IEEE,2008:413-420.
[36]Barnich O, Van Droogenbroeck M. ViBe:A universal background subtraction algorithm for video sequences[J]. Image Processing, IEEE Transactions on,2011, 20(6):1709-1724.
[37]Mahadevan V, Li W, Bhalodia V, et al. Anomaly detection in crowded scenes[C]//Computer Vision and Pattern Recognition (CVPR),2010 IEEE Conference on. IEEE,2010:1975-1981.
[38]Piciarelli C, Micheloni C, Foresti G L. Trajectory-based anomalous event detection[J]. Circuits and Systems for Video Technology, IEEE Transactions on, 2008,18(11):1544-1554.
[39]朱海龙,刘鹏,刘家锋,等.人群异常状态检测的图分析方法[J].自动化学报,2012,38(005)：742-750.
[40]许如初,宋恩民,陈卫东,等.覆盖平面上给定点集的最小凸多边形的算法[J].华中科技大学学报(自然科学版)1996.
[41]Albiol A, Silla M J, Albiol A, et al. Video analysis using corner motion statistics[C]//IEEE International Workshop on Performance Evaluation of Tracking and Surveillance.2009:31-38.