The Algorithm Research Of Moving Human Detection And Tracking Based On Video Sequence

With the increasing threat of terrorist, the advanced video surveillance system has to be put into use. The advanced video surveillance system needs to analyze the behaviors of people in order to prevent the occurrence of the potential dangerous case. The analysis of behaviors of people requires the human detection and tracking system. In recent years, the development of human detection and tracking system has been going forwards for several year, many real time systems have been developed. However, the actual monitoring results will be affected by many factors, thus automatic monitoring of the human body there are still many key technologies for urgent solution. This paper presented some algorithms research of moving human detection and tracking based on video sequence when the camera is fixed.In this paper, the author have researched and improved some related algorithm of human detection and tracking based on image processing and image analysis theory. Part of the moving human detection, some commonly used algorithms of moving human detection have been analyzed, and an effective detection method using vector dependence theory to resolve the disadvantage of images difference and inter-frame difference. The method calculated the vector matrix determinant WRONSKIAN of the reference frame and non-detection frame according to the same pixel linear correlation and the opposite non-linear nature of pixels. When the value was larger than the set threshold, the center vector would be determined as the blob of the moving object, else it was the blob of background. The result proved that this method was simple and effective than the images difference and inter-frame difference. This paper designed a shadow detection method to eliminate shape distortion due to shadow effect, and designed a processing procedure to eliminate noises.Part of moving human tracking, a target classification method based on codebook was presented to identify the human from other moving objects. This method used an algorithm for vector quantization based on classified and competition to design codebook. First, normalized the size of object, second, extract the shape of object as the features, and then matched the feature vector with the code vectors of codebook. The match process was to find a code vector in codebook with the minimum distortion to the feature vector of object. If the minimum distortion is less than a threshold, this object is human.In order to track the moving human more effective, a false objects detection method had been presented to detect the false objects caused by the following cases: such as sudden light change, humans in background removes and tree shaking. When there were integration and division of the crowd in the monitoring scenes, a"gray-scale statistical block"method have been presented to track the people before and after integration and division...