Use of a Novel Concept of Potential Pixel Energy for Object Tracking

In this paper, we propose a new method for kernel based object tracking which tracks the complete non rigid object. Definition the unio‎n image blob and mapping it to a new representation which we named as potential pixels matrix are the main part of tracking algorithm. The unio‎n image blob is constructed by expanding the previous object region based on the histogram feature. The potential pixels matrix of previous object and unio‎n blob are used for locating the target frame to frame. The formula for location of kernel in current frame is in the form of algebra and it does not need iterative process as mean shift tracker. For removing the background information and also the non rigid object changes, we propose to divide the expanding part of unio‎n blob to two parts and using distance threshold and similarity criterion for each part. We demonstrate the background removing of proposed method has better performance than the weighting kernel of mean shift tracking. Also the segmentation of object which produces narrow histogram is used for reduces the influence of background. We demonstrated the capability of the proposed method for several image sequences