A Complex Background Gesture Recognition Method

Since the publishment of "Computing machinery and intelligence", the research of artificial intelligence has experienced many stages, from expert systems to neural networks, and to statistical learning theory which is fashion today. A large number of excellent algorithms and theory have emerged during this period. At the same time, we should also realize that most of these methods are complex and hard to handle,especially not so capable in real system.This paper proposes an automatic method to recognize the hand gesture captured from video stream in a complex background. The method is just composed with some simple steps.The first step is to switch the pictures into YUV color space which has many outstanding attributes and will bring convenience in the further steps. Skin color detection is used to find out the probable hand region out of a complex background. Then a serial of morphological operations is performed to get the contour of the hand. At last the computation of Point-Curvature and Center-Angle is for marking out the fingertips. It is very useful to get the position of fingertips, as the meaning of hand gesture is always performed through fingertips and the movement of fingertips.Two simple systems have been implemented with the proposed algorithm, including a simple single finger tracking system and "stone scissors paper" game system. Although there are many problems existed in our method, it is still very constructive. And with the basis of this paper, we can implement more applications and achieve better results in the future.