Multi-Objects Segmentation and Tracking Based Graph-Cuts

In this paper we present an algorithm that joints segmentation and tracking of multiple objects in a video via graphcuts optimization technique. The proposed approach is composed of two steps. First, we initialize tracked objects through an initialization step based on background subtraction algorithm. Hence we obtain initial observations that will be used to predict the location of target object in the next frame. Then, we process a tracking step based on an energy function associated to each predicted observation. The minimization of this energy via graph-cut allows us to yield a better segmentation that matches extracted observations with initial detected objects. Experimental validation of the proposed method is performed in several video sequences and provides us significant tracking results.