XU Zhitong,LUO Yanmin,LIU Peizhong.Abnormal behavior detection of joint weighted reconstruction trajectory and histogram entropy[J].CAAI Transactions on Intelligent Systems,2018,13(06):1015-1026.[doi:10.11992/tis.201706070]





Abnormal behavior detection of joint weighted reconstruction trajectory and histogram entropy
徐志通12 骆炎民12 柳培忠3
1. 华侨大学 计算机科学与技术学院, 福建 厦门 361021;
2. 华侨大学 厦门市计算机视觉与模式识别重点实验室, 福建 厦门 361021;
3. 华侨大学 工学院, 福建 泉州 362021
XU Zhitong12 LUO Yanmin12 LIU Peizhong3
1. College of Computer Science and Technology, Huaqiao University, Xiamen 361021, China;
2. Xiamen Key Laboratory of Computer Vision and Pattern Recognition, Huaqiao University, Xiamen 361021, China;
3. College of Engineering, Huaqiao University, Quanzhou 362021, China
abnormal behavior detectionjointly weightedreconstruction trajectoryhistogram entropyKalman filtersHOG featuresspatio-temporal contextbehavior directory
To solve the problem of low precision in detecting multi-target abnormal behaviors such as fighting and robbery, in this paper, a novel human abnormal behavior detection algorithm is proposed based on a combination of the weighted reconstruction trajectory and histogram entropy. First, the background subtraction method is combined with aspect ratio to extract pedestrian targets, and then Kalman filters and histogram of oriented gradient (HOG) features are integrated into a spatio-temporal context algorithm to realize a robust tracking of completely occluded pedestrians in a short time, and finally the tracked trajectories are trained; a normal behavior directory is constructed for the motion trajectories defined as normal, and the trajectories to be tested are sparsely reconstructed. An effective detection of abnormal behavior is realized by combining the weighted minimum reconstruction residuals and amplitude direction histogram entropy. A comparative experiment shows that the algorithm can effectively detect abnormal behaviors such as fighting and robbery, and the detection rate can exceed 92% under static background without occlusions.


[1] CALDERARA S, HEINEMANN U, PRATI A, et al. Detecting anomalies in people’s trajectories using spectral graph analysis[J]. Computer vision and image understanding, 2011, 115(8):1099-1111.
[2] LI Weixin, MAHADEVAN V, VASCONCELOS N. Anomaly detection and localization in crowded scenes[J]. IEEE transactions on pattern analysis and machine intelligence, 2014, 36(1):18-32.
[3] 孙倩茹, 王文敏, 刘宏. 视频序列的人体运动描述方法综述[J]. 智能系统学报, 2013, 8(3):189-198 SUN Qianru, WANG Wenmin, LIU Hong. Study of human action representation in video sequences[J]. CAAI transactions on intelligent systems, 2013, 8(3):189-198
[4] VIRENDRA, SHETE V, UKUNDE N. Intelligent embedded video monitoring system for home surveillance[C]//Proceedings of 2016 International Conference on Inventive Computation Technologies. Coimbatore, India, 2016:1-4.
[5] JIANG Fan, YUAN Junsong, TSAFTARIS S A, et al. Anomalous video event detection using spatiotemporal context[J]. Computer vision and image understanding, 2011, 115(3):323-333.
[6] BOUTTEFROY P L M, BOUZERDOUM A, PHUNG S L, et al. Abnormal behavior detection using a multi-modal stochastic learning approach[C]//Proceedings of 2008 International Conference on Intelligent Sensors, Sensor Networks and Information Processing. Sydney, Australia, 2008:121-126.
[7] ZHAO Bin, LI Feifei, XING E P. Online detection of unusual events in videos via dynamic sparse coding[C]//Proceedings of 2011 IEEE Conference on Computer Vision and Pattern Recognition. Providence, RI, USA, 2011:3313-3320.
[8] LU Cewu, SHI Jianping, JIA Jiaya. Abnormal event detection at 150 FPS in MATLAB[C]//Proceedings of 2013 IEEE International Conference on Computer Vision. Sydney, NSW, Australia, 2013:2720-2727.
[9] LI Ce, HAN Zhenjun, YE Qixiang, et al. Abnormal behavior detection via sparse reconstruction analysis of trajectory[C]//Proceedings of the 6th International Conference on Image and Graphics. Hefei, Anhui, China, 2011:807-810.
[10] 李海霞, 范红. 基于背景差法的几种背景建模方法的研究[J]. 工业控制计算机, 2012, 25(7):62-64 LI Haixia, FAN Hong. Research of several background modeling based on background subtraction[J]. Industrial control computer, 2012, 25(7):62-64
[11] GOYAL K, SINGHAI J. Review of background subtraction methods using Gaussian mixture model for video surveillance systems[J]. Artificial intelligence review, 2017.
[12] LI Yanshan, HUANG Qinghua, XIE Weixin, et al. A novel visual codebook model based on fuzzy geometry for large-scale image classification[J]. Pattern recognition, 2015, 48(10):3125-3134.
[13] 傅博, 李文辉, 陈博, 等. 基于加权光流能量的异常行为检测[J]. 吉林大学学报:工学版, 2013, 43(6):1644-1649 FU Bo, LI Wenhui, CHEN Bo, et al. Abnormal behavior detection based on weighted energy of optical flow[J]. Journal of Jilin university:engineering and technology edition, 2013, 43(6):1644-1649
[14] APEWOKIN S, VALENTINE B, FORSTHOEFEL D, et al. Embedded real-time surveillance using multimodal mean background modeling[M]//KISA?ANIN B, BHATTACHARYYA S S, CHAI S. Embedded Computer Vision. London, UK:Springer, 2009:163-175.
[15] 叶锋, 范曼曼, 郑子华, 等. 一种改进的基于平均背景模型的运动目标检测算法[J]. 福建师范大学学报:自然科学版, 2011, 27(4):44-49 YE Feng, FAN Manman, ZHENG Zihua, et al. An improvement of moving object detection algorithm based on average background model[J]. Journal of Fujian normal university:natural science edition, 2011, 27(4):44-49
[16] 李庆武, 蔡艳梅, 徐立中. 基于分块分类的智能视频监控背景更新算法[J]. 智能系统学报, 2010, 5(3):272-276 LI Qingwu, CAI Yanmei, XU Lizhong. Background update algorithm based on blocks classification for intelligent video surveillance[J]. CAAI transactions on intelligent systems, 2010, 5(3):272-276
[17] ZHANG Kaihua, ZHANG Lei, LIU Qingshan, et al. Fast visual tracking via dense Spatio-temporal context learning[C]//Proceedings of the 13th European Conference on Computer Vision. Zurich, Switzerland, 2014:127-141.
[18] 汤春明, 卢永伟. 基于改进的稀疏重构算法的行人异常行为分析[J]. 计算机工程与应用, 2017, 53(8):165-169 TANG Chunming, LU Yongwei. Pedestrian abnormal behavior analysis based on optimized sparse reconstruction algorithm[J]. Computer engineering and applications, 2017, 53(8):165-169
[19] WANG Zhenhai, XU Bo. An effective object tracking based on spatio-temporal context learning and Hog[C]//Proceedings of the 201511th International Conference on Natural Computation. Zhangjiajie, China, 2015:661-664.
[20] LI Ce, HAN Zhenjun, YE Qixiang, et al. Visual abnormal behavior detection based on trajectory sparse reconstruction analysis[J]. Neurocomputing, 2013, 119:94-100.
[21] 杜鉴豪, 许力. 基于区域光流特征的异常行为检测[J]. 浙江大学学报:工学版, 2011, 45(7):1161-1166 DU Jianhao, XU Li. Abnormal behavior detection based on regional optical flow[J]. Journal of Zhejiang university:engineering science, 2011, 45(7):1161-1166
[22] FRANKLIN J. The elements of statistical learning:data mining, inference and prediction[J]. The mathematical intelligencer, 2005, 27(2):83-85.
[23] ISCEN A, WANG Yijie, DUYGULU P, et al. Snippet based trajectory statistics histograms for assistive technologies[C]//Proceedings of European Conference on Computer Vision. Zurich, Switzerland, 2014:3-16.
[24] 周同驰, 程旭, 吴镇扬. 分层树结构字典编码的行为识别[J]. 中国图象图形学报, 2014, 19(7):1054-1061 ZHOU Tongchi, CHENG Xu, WU Zhenyang. Action recognition using hierarchically tree-structured dictionary encoding[J]. Journal of image and graphics, 2014, 19(7):1054-1061
[25] SCHÖLKOPF B, PLATT J, HOFMANN T. Efficient sparse coding algorithms[C]//Proceedings of 2006 Conference on Advances in Neural Information Processing Systems.[s.l.]2007:801-808.
[26] CONG Yang, YUAN Junsong, LIU Ji. Sparse reconstruction cost for abnormal event detection[C]//Proceedings of 2011 IEEE Conference on Computer Vision and Pattern Recognition. Providence, RI, USA, 2011:3449-3456.
[27] 杨玉梅. 基于信息熵改进的K-means动态聚类算法[J]. 重庆邮电大学学报:自然科学版, 2016, 28(2):254-259 YANG Yumei. Improved K-means dynamic clustering algorithm based on information entropy[J]. Journal of Chongqing university of posts and telecommunications:natural science edition, 2016, 28(2):254-259
[28] 刘燕, 高云. 利用角点历史信息的异常行为识别算法[J]. 计算机工程与科学, 2014, 36(6):1127-1131 LIU Yan, GAO Yun. Abnormal behavior recognition based on corner motion history[J]. Computer engineering and science, 2014, 36(6):1127-1131
[29] WANG Lijun, DONG Ming. Detection of abnormal human behavior using a matrix approximation-based approach[C]//Proceedings of the 201413th International Conference on Machine Learning and Applications. Detroit, MI, USA, 2014:324-329.
[30] 林玲, 廖德, 高阳, 等. 基于加权样本选择与主动学习的视频异常行为检测算法[J]. 模式识别与人工智能, 2016, 29(4):341-349 LIN Ling, LIAO De, GAO Yang, et al. Video anomaly detection algorithm based on weighted sample selection and active learning[J]. Pattern recognition and artificial intelligence, 2016, 29(4):341-349


更新日期/Last Update: 2018-12-25