Based On Video Object Detection And Tracking Of Research

With the development of science and technology, the intelligent video surveillance system, which has video analysis function and detect and tract the moving object, has become the hot point in the research. Detection and tracking of moving object in image sequence is an important area in the domain of digital image processing, patern recognition and computer vision. Its applications include that robot navigation, security surveillance, medical image analysis and virtual reality.For the probem of the detection and tracking of the given object, this paper studies the specific target detection and tracking theories in motionless background. Specifically says:First of all,it uses the color segmentation of background subtraction method to detect moving objects, and then get the binary, morphology filtering and connectivity detection methods to locate the target, the method can more accurately be fix the positon of targets.It can avoid the interference of background to the track.The second,it uses the color recognition to determine the face of the Specific region and adopt improved methods of adaboost methods and haar-type characteristics to determine the face region.Then it ultimately determine the location of the face.the third,in target tracking context,this article discussed CamShift (Continuous Adaptive Meanshift) tracking algorithm.As the object of the color information is the inherent characteristics of the object,the color characteristics of the moving object on the track can solve the block of the target after the loss. CamShift tracking algorithm is based on the tracking method of the target color characteristics. because of its nothing parameters, fast speeding computing have been widely used. This issue combine the results of the face detection and camShift tracking algorithm to improve the speed of the match tracking.it achieve the goal of the fast track. In this paper, the experimental procedure use the Visual C++to achieve.