DETECTION OF CROWD BEHAVIOR FROMLIVE STREAMING

Abstract

“Crowd Behavior Detection From Live Streaming”, is a surveillance project which observes crowd optical flow and detect any abnormal events in crowds. A proactive approach is required to effectively manage the crowd flow and to accurately detect the erratic behavior of crowd. So, we have proposed an algorithm to detect the behavior of crowd. Frames are taken at regular intervals through a video camera and these frames are processed and classified further to detect the normal or abnormal activities of crowd. A novel motion vector based technique is used to detect the cluster of interest. The features of the motion vectors are analyzed to characterize the crowd behavior. The results on simulated crowds demonstrate the effectiveness of the proposed algorithm.