Eye-hand Hybrid Gesture Recognition System for Human Machine Interface

Gesture Recognition has become a way for computers to recognise and understand human body language. They bridge the gap between machines and human beings and make the primitive interfaces like keyboards and mice redundant. This paper suggests a hybrid gesture recognition system for computer interface and wireless robot control. The real-time eye-hand gesture recognition system can be used for computer drawing, navigating cursors and simulating mouse clicks, playing games, controlling a wireless robot with commands and more. The robot illustrated in this paper is controlled by RF module. Playing a PING-PONG game has also been demonstrated using the gestures. The Haar cascade classifiers and template matching are used to detect eye gestures and convex hull for finding the defects and counting the number of fingers in the given region. Keyword – hybrid gesture, machine interaction, real-time vision system, eye gesture recognition.