Abnormal Behavior Detection Based on Global Motion Orientation

摘要：

　　A novel approach is introduced in this paper to detect abnormal behavior based on global motion orientation.Compare to the normal behavior (walking,shaking hands etc.),abnormal behavior has different orientation.The method we introduced divides each frame into blocks,makes statistical analysis of the global motion direction histogram of all frame blocks and extracts characteristics.At last,behavior is detected with support vector machine (SVM).Experiment shows that the method proposed in the paper has certain robustness and can achieve real-time monitoring.

关键词： abnormal behavior detection global motion orientation support vector machine (SVM)

作者: Xuyan Ma Guomao Liang Wei Yu Zhiyi Qu

作者单位: School of Information Science & Engineering, Lanzhou University, Lanzhou, China

会议类型: 国际会议

会议名称: 2013 2nd International Conference on Systems Engineering and Modeling(ICSEM-13)(2013年第二届系统工程与建模国际会议)

会议地点: 北京

会议语种:英文

页码: 461-464

在线出版日期: 2013-04-19（万方平台首次上网日期，不代表论文的发表时间）