Abnormal behavior detection by causality analysis and sparse reconstruction

Abstract

A new approach for abnormal behavior detection was proposed using causality analysis and sparse reconstruction. To effectively represent multiple-object behavior, low level visual features and causality features were adopted. The low level visual features, which included trajectory shape descriptor, speeded up robust features and histograms of optical flow, were used to describe properties of individual behavior, and causality features obtained by causality analysis were introduced to depict the interaction information among a set of objects. In order to cope with feature noisy and uncertainty, a method for multiple-object anomaly detection was presented via a sparse reconstruction. The abnormality of the testing sample was decided by the sparse reconstruction cost from an atomically learned dictionary. Experiment results show the effectiveness of the proposed method in comparison with other state-of-the-art methods on the public databases for abnormal behavior detection.

Keywords

abnormal behavior detection / Granger causality test / causality feature / sparse reconstruction

References

[1] CuiX, LiuQ, GaoM, MetaxasD N. Abnormal detection using interaction energy potentials [C]. IEEE Computer Vision and Pattern Recognition. Colorado Springs: IEEE, 201131613167

[2] PiciarelliC, ForestiG L. Trajectory-based anomalous event detection [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2008, 18(11): 1544-1554

[3] DeepikaR, PrasathV S A, IndhumathiM, KumarP M A, VaidehiV. Anomalous event detection in traffic video surveillance based on temporal pattern analysis [J]. Journal of computers, 2017, 12(2): 190-199

[4] KwakS, ByunH. Detection of dominant flow and abnormal events in surveillance video [J]. Optical Engineering, 2011

[5] MahadevanV, LiW, BhalodiaV, VasconcelosN. Anomaly detection in crowded scenes [C]. IEEE Computer Vision and Pattern Recognition, 2010, San Francisco, IEEE: 19751981

[6] WuS, MooreB E, ShahM. Chaotic invariants of Lagrangian particle trajectories for anomaly detection in crowded scenes [C]. IEEE Computer Vision and Pattern Recognition, 2010, San Francisco, IEEE: 20542060

[7] KratzL, NishinoK. Anomaly detection in extremely crowded scenes using spatio-temporal motion pattern model [C]. IEEE Computer Vision and Pattern Recognition, 2009, Miami Beach, IEEE: 14461453

[8] MehranR, OyamaA, ShahM. Abnormal crowd behavior detection using social force model [C]. IEEE Computer Vision and Pattern Recognition, 2009, Miami Beach, IEEE: 935942

[9] BenezethY, JodinP, SaligramaV, RosenbergerC. Abnormal events detection based on spatio-temporal co-occurences [C]. IEEE Computer Vision and Pattern Recognition, 2009, Miami Beach, IEEE: 24582465

[10] CongY, YuanJ S, LiuJ. Sparse reconstruction cost for abnormal event detection [C]. IEEE Computer Vision and Pattern Recognition, 2011, Colorado Springs, IEEE: 34493456

[11] KimJ, GraumanK. Observe locally, infer globally: A space-time MRF for detecting abnormal activities with incremental updates [C]. IEEE Computer Vision and Pattern Recognition, 2009, Miami Beach, IEEE: 29132920

[12] SaligramaV, ChenZ. Video anomaly detection based on local statistical aggregates [C]. IEEE Computer Vision and Pattern Recognition, 2012, Providence, IEEE: 21122119

[13] ReddyV, SandersonC, LovellB C. Improved anomaly detection in crowded scenes via cell-based analysis of foreground speed, size and texture [C]. IEEE Computer Vision and Pattern Recognition, 2011, Colorado Springs, IEEE: 5561

[14] NallaivarothayanH, FookesC, DenmanS. An MRF based abnormal event detection approach using motion and appearance features [C]. IEEE International Conference on Advanced Video and Signal based Surveillance, 2014, Seoul, IEEE: 343348

[15] FuZ Y, HuW M, TanT I. Similarity based vehicle trajectory clustering and anomaly detection [C]. IEEE International Conference on Image Processing, 2005, Genoa, IEEE: 20292032

[16] HuW M, XiaoX, FuZ Y. A system for learning statistical motion patterns [J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2010, 28(9): 1450-1464

[17] BeraA T, KimS, ManochaD. Realtime anomaly detection using trajectory-level crowd behavior learning [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2016, Las Vegas, IEEE: 5057

[18] JeongH, ChangH J, JinY C. Modeling of moving object trajectory by spatio-temporal learning for abnormal behavior detection [C]. IEEE International Conference on Advanced Video and Signal-Based Surveillance, 2011, Klagenfurt, IEEE: 119123

[19] YangW Q, GaoY, CaoL B. TRASMIL: A local anomaly detection framework based on trajectory segmentation and multi-instance learning [J]. Computer Vision and Image Understanding, 2013, 117(10): 1273-1286

[20] LiC, HanZ, YeQ, JiaoJ. Visual abnormal behavior detection based on trajectory sparse reconstruction analysis [J]. Neurocomputing, 2013, 119(7): 94-100

[21] ZhaoB, LiF F, XingE. Online detection of unusual events in videos via dynamic sparse coding [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2011, Colorado Springs, IEEE: 33133320

[22] RenH M, LiuW F, EscaleraS, MoeslundT B. Unsupervised behavior-specific dictionary learning for abnormal event detection [C]. The British Machine Vision Conference, 2015, Swansea Bay, Springer: 135144

[23] WangX, MaX, GrimsonW. Unsupervised activity perception in crowded and complicated scenes using hierarchical Bayesian models [J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(3): 539-555

[24] VaswaniN, Roy-ChowdhuryA, ChellappaR. Shape activity: A continuous state hmm for moving/deforming shapes with application to abnormal activity detection [J]. IEEE Transactions on Image Processing, 2005, 14(10): 1603-1616

[25] MoX, MongaV, BalaR, FanZ G. Adaptive sparse representations for video anomaly detection [J]. IEEE Transactions on Circuits and Systems for Video Technology, 2014, 24(4): 631-645

[26] HanM, XuW, TaoH. An algorithm for multiple object trajectory tracking [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2004, Washington, IEEE: 864871

[27] CuiX Y, LiuQ S, GaoM C, MetaxasD N. Abnormal detection using interaction energy potentials [C]. IEEE Conference on Computer Vision and Pattern Recognition, 2011, Colorado Springs, IEEE: 31613167

[28] XiaL M, YangB J, TuH B. Recognition of suspicious behavior using case-based reasoning [J]. Journal of Central South University, 2015, 22(1): 241-250

[29] GrangerC. Investigating causal relations by econometric models and cross-spectral methods [J]. Econometrica, 1969, 37(3): 424-438

[30] XiaL M, HanF, WangJ. Complex human activities recognition using interval temporal syntactic model [J]. Journal of Central South University, 2016, 23(10): 2578-2586

[31] ShiJ, TomasiC. Good features to track [C]. IEEE Conference on Computer Vision and Pattern Recognition, 1994, Seattle, IEEE: 593600

[32] LaptevI. On space-time interest points [J]. International Journal of Computer Vision, 2005, 1(23): 432-439

[33] ZhouY, YanS, HuangT. Pair-activity classification by bitrajectory analysis [C]. IEEE Conference on Computer Vision and Pattern Recognition, 200836823689

[34] NiB P, YanSC, KassimA. Recognizing human group activities with localized causalities [C]. IEEE Conference on Computer Vision and Pattern Recognition, 200914701477

[35] NesterovY. Gradient methods for minimizing composite objective function [J]. Mathematical Programming, 2013, 140(1): 125-161