Decomposition approach for hand-pose recognition

Abstract

This paper presents the recognition of hand-pose by decomposition approach, which is constructed on the hierarchical structure of hand-pose classifiers, and subgroup classifier. With the new method, it is possible to do hand-pose specified recognition with simple structure. To show the effectiveness of the proposed system, experiments were carried out with 32 hand-poses, grouped into 12 subgroups and demonstrated a 96.46% success rate for individual poses.