Real-Time Detection and Understanding of Isolated Protruded Fingers

Making sense of hand gestures in applications such as novel user interfaces is an active area of research. Many current methods use complex 3D model of non-rigid hands and involved paradigms in solving the general problem. In this paper we look at a subproblem of sign language alphabet recognition where gestures are made with protruded fingers. In contrast to the more intricate schemes, we propose a simple, fast classification algorithm using two dimensional projection of Euler angles. Not only is our scheme novel, it lends itself to a real time implementation.