Pedestrian violence detection based on optical flow energy characteristics

Pedestrian violence detection has important application value and research significance in many fields, such as security of society, public monitoring, and personal safety. The traditional behavior detection methods based on analyzing the structure and characteristics of human beings, can distinguish change of human behavior over of time. However, the high computational complexity and the limitations of the study model make it unsuitable for realtime detection. To overcome the drawbacks of the existing behavior detection methods, this paper presents a new method of pedestrian violence detection based on optical flow energy characteristics. In the proposed method, firstly, the corner joints of pictures are detected using the Shi-Tomasi corner detection algorithm. Next, the optical flow parameter of moving objects is calculated using an improved Lucas-Kanade pyramid optical flow algorithm. Finally, violent behavior is detected using the histogram of the computed optical flow energy values. The illustrative results show that the method presented in this paper can be used for real-time detection of violence in public.

[1]  Shaogang Gong,et al.  Incremental and adaptive abnormal behaviour detection , 2008, Comput. Vis. Image Underst..

[2]  Christopher G. Harris,et al.  A Combined Corner and Edge Detector , 1988, Alvey Vision Conference.

[3]  Daniel P. Huttenlocher,et al.  Pictorial Structures for Object Recognition , 2004, International Journal of Computer Vision.

[4]  Carlo Tomasi,et al.  Good features to track , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[5]  Mubarak Shah,et al.  Person-on-person violence detection in video data , 2002, Object recognition supported by user interaction for service robots.

[6]  Yangsheng Xu,et al.  Abnormal Behavior Detection by Multi-SVM-Based Bayesian Network , 2007, 2007 International Conference on Information Acquisition.

[7]  Richard P. Kleihorst,et al.  Abnormal motion detection in a real-time smart camera system , 2009, 2009 Third ACM/IEEE International Conference on Distributed Smart Cameras (ICDSC).

[8]  Greg Mori,et al.  Action recognition by learning mid-level motion features , 2008, 2008 IEEE Conference on Computer Vision and Pattern Recognition.

[9]  A. Beghdadi,et al.  Abnormal behavior detection using a multi-modal stochastic learning approach , 2008, 2008 International Conference on Intelligent Sensors, Sensor Networks and Information Processing.