Research On Image Storage Of Video Surveillance System Based On Image Analyzing

Video surveillance is one of the application technology focuses which people have paid close attention to. Video surveillance has been used in a lot of occasions because of its visibility, convenience and abundant in content. To a set of video surveillance system, a big chunk of function is collecting evidence afterwards. The traditional video surveillance system is to carry on the all-weather video to the scene. This method not only wasted a large amount of storage medium but also bring the inconvenience to search of information while collecting evidence afterwards. In this paper a method based on image analyzing has been proposed to judge whether to record image or not. In this way, we can reduce recording amount of the image without value.Image analyzing is the key to realizing the information alternative of the image storage. This paper realize this alternative mechanism by analyzing moving object in the image sequences. This paper has introduced the relevant theory of image processing and moving object segmentation at first, then adopt the background subtraction to realize the moving object segmentation in the image, and analyze the state of moving object through setting up distinguishing matrix and state recognition algorithm. At last we design the image-recording rule according to these. The hardware system in this paper is a system based on DSP. In order to realize local image storage, image compression system and USB host system based on DSP has been designed in this paper.The system this paper proposed reduced the quantity of image storage while guaranteeing to record sensitive scene, have offered a new solution for the image storage of video surveillance system.