Face Tracking Based on Camshift Algorithm and Motion Prediction

In the face tracking process, if occlusion occurred, the search window of Camshift tracker would converge to a minimum and remain in a local area, thereby causing tracking failure. In this paper, we proposed a new tracking algorithm to overcome this problem by integrating the Camshift algorithm and alpha-beta-gamma filtering prediction. After obtaining the human face location in the first three frames by Camshift, we switch to initializing alpha-beta-gamma filter parameters, and estimate reference point of the face candidate in the next frame using alpha-beta-gamma filtering algorithm. In the absence of occlusion, the predicted point can be used as the initial iteration point by Camshift algorithm to reduce the number of iterations and achieve real-time performance; otherwise, the value obtained by alpha-beta-gamma filtering prediction will be used to replace the Camshift output to ensure the continuity of tracking. Experimental results show that the proposed algorithm is effective and robust.