Robust hand posture recognition using multi-feature for robot vision

This paper introduces a novel approach for hand posture recognition based on multi-feature fusion. We cast hand posture recognition as a sparse representation problem. By integrating different features, including color, texture and shape feature, the proposed method can take advantage of each feature and hence is robust to partial occlusion and varying illumination. With the proposed method, we developed an application for our intelligence service robot, in which it demonstrates the effective and robust performance.