本田 雄一郎

In this article a new pattern recognition system for hand gestures developed for the control of active hand prostheses is presented. The system uses muscle contraction for gesture recognition. Different hand gestures produce discrete tactile patterns in the prosthetic socket by muscle contractions. The tactile patterns from hand gestures are analyzed and classified by the recognition system. Thus, a user can control hand prosthesis muscle contraction of the arm from handgestures. This paper introduces the concept of force resistive resistor sensor based recognition system for hand gestures and then discusses the issues of stability of classification during reattaching of the prosthetic socket.