Hand Gesture Tracking using Particle Filter with Multiple Features

Hand gesture tracking is one of core method in the field of Human Computer Interaction. In this paper we proposed a tracking method based on Particle Filter framework. Particle Filter algorithm provides a powerful open framework for target tracking and can efficiently deal with motion estimation under the non-linear and nonGaussian environments. The novelty of our method is in the combination of color cue and LBP cue for the representation of the target hand under the particle filter framework. The LBP cue mainly focus on the detail information of the hand gesture, which can effectively cope with clutters due to possible skin-colored objects from the background. Experimental result shows that the proposed tracking method is robust and could track hand gesture satisfactorily.