Occlusion Target Tracking

In recent years, with the improving requirements from the military and others, there are a great of research interests in studying of tracking technology of moving target from image sequence in complex background.The essence of object tracking is automatic recognizing moving target, deciding the state and the position of the target, and automatic tracking the target. It is difficult because of the unstable image signal and complicated environmental factors. Especially, occlusion problem gradually became the limitation of the practical tracking algorithm.In the dissertation, many important results which bring from new ideas are achieved and listed as the following:1. For the tracking work, based on moving target detecting, in this method, first, make template, then find a match template with multiple resolutions. The intensity correlation matching based on multi-block method can estimate the occlusion region accurately through the blocks with distinct feature, and track the target by the remaining unoccluded block.And update the template if necessary, while predict the moving parameters through karlman filter. So the problems of keep-out target, sensitive detailed pixels and computational complexity are solved.2. The object extraetion based on the moving area is studied. The edge is the feature for rigid objects (the edges are used to the edge matching tracking algorithm). And then,use spreading feature algorithm to detect the feature points.Then the points vote the realative shift.3. Based on in depth studying on extraction of features and comparability matching algorithm on the basis of HIS colors accumulative histogram, which is in accordance with visual idiosyncrasy of human beings. First, we use karlman filter to predict the moving parameters. Then we ues algorithm on the basis of HIS colors accumulative histogram to accumulative the relative shift. So the problems of keep-out target, sensitive detailed pixels and computational complexity are solved.