Non-rigid object tracking

In this paper, we propose a new algorithm for optimally adapting ellipses outlining objects of interest in order to improve the performance of a colour based tracking approach for real video sequences. We present a Lagrangian based method to integrate a regularising component into the covariance matrix to be computed. Technically, we intend to reduce the residuals between the estimated probability distribution and the expected one. We argue that by doing so, the shape of the ellipse can be properly adapted in the tracking stage. Experimental results confirm that our proposed method leads to favourable performance in terms of shape adaption and object localisation.