A novel approach for human-machine interaction using hand gesture

Abstract

Human machine interaction is a rapidly developing field. Hand gestures are used to interact with the machine; this makes the human machine interaction very interesting and intuitive. Several devices have been developed in the past using the camera for recognizing hand gestures and communicating with the machine. In this paper we have proposed a new system called hand gesture recognition (HGR) to overcome the use of camera which has several disadvantages like lighting, making proper hand movements before camera, operating in less range. The universal control unit may be as a form of wearable one. A Kalman filtering approach is proposed for stable displacement estimation and waveband classification of the measured acceleration. With the sensors placed in the form of watch or band on our hand, any motion made by the hand will be detected by these sensors and relayed to a control unit to take the required actions. The estimated results are compared with those of a commercial geometry measurement.