Research On Synergistic Tracking Based On Homography Constraint

With the development of society,social security has become a topic of growing concern.Monitoring and tracking based on machine vision has been widely used in the public domain.Due to the explosive growth of information,the need for on-site monitoring of the human needs of high cost,so the video surveillance system to be able to actively assist in monitoring,reduce labor.Traditional single camera target tracking accuracy is low.In some complex scenes,it is difficult to achieve tracking requirements for single camera target tracking.The amount of information acquired by multiple cameras is better than that of single camera.Multi view information fusion can be achieved by homography constraint,and multi camera tracking can overcome some problems that can not be solved by single camera.A new background updating algorithm is adopted in the target detection module.It is a new target detection algorithm based on frame difference method,background difference method and image block classification.In the aspect of target localization,In this paper,we use the homography matrix to obtain the location of the target foot region.A rectangular box is drawn from the angle of tolerance of the tolerance error to estimate the foot area on the view plane.Then,the gray histogram of the foot estimation region is estimated,and then the speckle classification is performed by comparing the similarity of gray histogram.In the aspect of target tracking,this paper uses the Euclidean distance algorithm of the foot spot centroid to improve the real-time and accuracy of tracking.Through the simulation experiment,the target detection in this paper is more complete than the traditional method.Target location and tracking results are reliable.The real-time performance of the whole algorithm is better.