Posterior Probability Object Tracking Method Using Momentum Based Level Set

This paper proposes a novel object tracking method that is robust to a cluttered background and a large motion. First, a posterior probability measure (PPM) is adopted to locate the object region. Then the momentum based level set is used to evolve the object contour in order to improve the tracking precision. To achieve rough object localization, the initial target position is predicted and evaluated by the Kalman filter and the PPM, respectively. In the contour evolution stage, the active contour is evolved on the basis of an object feature image. This method can acquire more accurate target template as well as target center. The comparison between our method and the kernel-based method demonstrates that our method can effectively cope with the deformation of object contour and the influence of the complex background when similar colors exist nearby. Experimental results show that our method has higher tracking precision.