ABSTRACT
This paper presents a way of recognizing hand gestures that can be detected quickly and accurately, using techniques of Motion Detection and Skin Detection eliminating pixels that are not important in the identification of the coordinate gestures and to leave only pixels that bring a standard tone of human skin to the Haar Cascade for gesture recognition. Besides that, the new algorithm also delivers to CamShift an image that is faster tracked than the one that is produces using Haar Cascade only.
