Cursor Movement in Real Time using Hand Gesture Recognition.

This paper proposes a method for cursor movements by Classifiers. Purpose of this paper is to reduce the external hardware and to interact with the computer. The system consists of two modules namely, gesture recognition and control of cursor movements. Gesture attained using neural net training to the cascade Haar classifiers. Supervised neural net training helps in removing non-gesture pattern that helps to classify a frame as a gesture. The frame from the video recorded using web cam is given as input. Based on the extracted features, from the frame the cursor movements are controlled. This method improves the computational speed and also the effectiveness of communication. The tool used here is OpenCV 1.0 under Visual Studio 2008.