Intelligent Video Monitoring System Based On Dvr Software Design

Though a lot of drug rehabilitation centers have realized the traditional video surveillance system, they usually used to collect evidence for the future and they did not give full play to the initiative and intelligence of the video surveillance system. According to the actual situation of video surveillance of drug rehabilitation in YunNan Province, the main purpose of this paper is to do some research and system development on the key technology of video surveillance of drug rehabilitation which is based on behavior pattern recognition.Based on the theory and technology of digital image processing, we extracted and tracked the objects appeared in the scene. Through the analysis of the moving objects’ geometric features, we have defined the alter rules of abnormal behavior when they occurs, the abnormal behavior include Intrusion, over the window, over the wall and assemble, and finally realized the design of video surveillance system which is based on DVR. The main contributions of this paper are as follows:1. Analyzed and compared the advantages and disadvantages of the common objects detection algorithms, based on the surendra background updating algorithm and adaptive threshold method, we proposed an improved surendra background updating algorithm to update background and extract the moving objects. Experiments show that the improved algorithm not only solves the problem of mistaking the objects for background when there are moving objects in the first frame of the video sequence, but also can extract the objects from background adaptively.2. After studying of the objects tracking algorithm based on blobs and kalman algorithm deeply, we used an algorithm which combined the blob and kalman filter algorithm together for object tracking at last. Experimental results showed that the algorithm can reduce the calculation greatly, so that can satisfy the need of real-time of Video Surveillance System.3. In the abnormal behavior detection module, due to the geometric feature which will change when all kinds of abnormal occurs, we defined alert rules for all these abnormal behaviors. The experimental results showed that the alert rules defined are applicable when they occur.4. Combined the modules of objects detection, objects tracking and abnormal detection, we designed and realized an intelligent video surveillance system.