Video Analysis and Study of Abnormal Behavior Based on Subway Complicated Scene
Zhang Qigu · Video Engineering · 2014
In order to reduce potential safety hazard which is caused by abnormal behavior in the field of public surveillance,from the perspective of artificial intelligence,the complicated subway is as a specific research background. The characteristics of the abnormal behavior are analyzed,and the system is designed. Video surveillance images are calibrated quickly,and the scene model and abnormal individual behavior model is built based on the context. The algorithm of improved camshaft is used to fill the track continuously. Abnormal individual behavior is characterized with feature points in the trajectory set of properties. The experimental results illustrate that it can analyze real time of abnormal behavior based on subway scene,achieve the goal of identify abnormal behavior rapidly,and the detection rate is 89. 9%.