Real time hand tracking by combining particle filtering and mean shift

Abstract

Particle filter and mean shift are two successful approaches taken in the pursuit of robust tracking. Both of them have their respective strengths and weaknesses. In this paper, we proposed a new tracking algorithm, the mean shift embedded particle filter (MSEPF), to integrate advantages of the two methods. Compared with the conventional particle filter, the MSEPF leads to more efficient sampling by shifting samples to their neighboring modes, overcoming the degeneracy problem, and requires fewer particles to maintain multiple hypotheses, resulting in low computational cost. When applied to hand tracking, the MSEPF tracks hand in real time, saving much time for later gesture recognition, and it is robust to the hand's rapid movement and various kinds of distractors.