Methods of global optimization in the tracking of contours

A new method for tracking contours of moving objects in clutter is presented. For a given object, a model of its contours is learned from training contours in the form of a subset of curve space. Complexity is added to the contour model by analyzing rigid and non-rigid transformations of contours separately. In the course of tracking, a very large number of potential curves are typically observed due to the presence of extraneous edges in the form of clutter; the learned model guides the algorithm in picking out the correct one. The algorithm is posed as a solution to a minimization problem; theoretical results on how to achieve the global minimum to within a certain resolution, and the complexity of this operation, are presented. Experimental results from applying the proposed algorithm to the tracking of a flexing finger and to a conversing individual´s lips are also presented.