Finger Parts Based Hand Gesture Recognition

Hand gesture is an interface between human and computer. It allows complex operation using hand gesture movement. Aim of this paper is to develop a scheme that recognizes bare hand gesture and perform certain actions depending on finger count. Bare hand gesture is used for recognition instead of using color glove or data glove. We are controlling the winamp player using finger count and angle. For gesture recognition we are performing noise removal, thresholding. Thresholding is occurs on the basis of skin tone selection. So depending on skin color object image is black and other background image is white. After that using circular profiling architecture we are recognizing the gestures. Proposed scheme can effectively reduce the response time and improve the accuracy.