A novel quadrilateral-based tracking method

This paper proposes a novel tracking method based on quadrilateral-based segmentation. The tracking method matches quadrilaterals in a region with those in a reference frame to construct region correspondences, from which trajectory of each region can be obtained. The Table Tennis sequence has been used to evaluate the tracking method. Experiment results show that the proposed tracking method can track the ping-pong ball and the racket reasonably well over a series of images, indicating matching quadrilaterals to track region is a viable approach.