| With the rapid development of information technology in today’s society, intelligent video surveillance has been widely used in all fields of life. Analysis of human movement abnormal behavior is an important research direction of intelligent video surveillance. It uses digital image processing technology for detecting and tracking human in video sequences. And on this basis to analyze and judge human behavior, it can timely alarm when abnormal behavior happened. Intelligent video surveillance to detect abnormal behavior of human movement saves a lot of manpower and material resources. It can maximum reduce the rate of false positives and false negatives.This paper analyzes the behavior of human movement abnormalities, specifically from the following three aspects do the research:Frist, Moving object detection. In this paper, inter-frame difference and background subtraction method are combined. A method of updating the background is put forward. This paper discusses the post-processing method of extracted moving object, such as region growing, shadow elimination; Second, Moving object tracking. This paper uses the area tracking method based on feature matching, and uses Kalman filter to predict the position of the moving object. Object area of overlap, color histogram and predicting position is selected as the featurs to match the foreground block and object; Finally, abnormal human behavior analysis. The paper uses analysis method based on behavioral characteristics. By extracting the object behavior characteristics, such as moving trajectory, speed, relevance, the paper has realized the detection of abnormal behavior. The abnormal behaviors include running, falling, fighting and wandering are analyzed in the paper.A large number of experiments show that this approach has a good adaptability and low algorithm complexity and so on. The scheme is able to detect the abnormal human behavior. It can be widely used in residential area, square, subway and so on. |
PDF Full Text Request