Dynamic Hand Gesture Recognition Using Kinematic Features Based on Improved Dynamic Time Warping ?

The use of hand gesture provides an attractive alternative to cumbersome interface devices for human computer interface. In our research, we exploit multiple cues including divergence features, vorticity features and hand motion direction vector. Divergence and vorticity are derived from the optical flow for hand gesture recognition in videos. The hand tracking algorithm finds the centroids of the moving hand regions, computes hand motion direction vector. For efficient recognition of gestures made at varying rates, an improved dynamic time warping method be introduced to our experiments where we have 12 isolated gestures and a library of 300 gestures videos. Those experimental results demonstrate that the proposed approach yields a satisfactory and higher recognition rate.