HAND GESTURE RECOGNITION BY USING NEURAL NETWORK

摘要

Hand Gesture is consistently utilized in daily life style. It is so ordinary technique to communicate. Hand gesture recognition technique is extensively used in the application area of managing mouse and/or keyboard functionality, 3D World, mechanical system, Manipulate virtual objects, Human/Robot Manipulation Navigate in a Virtual Environment, and Instruction Communicate at a remote location. This paper initiates a real time hand gesture recognition system. The proposed system include of four phases: image acquisition, feature extraction, and recognition, text to voice conversion. In the image acquisition phase input image of hand gestures are taken by digital camera or web camera in fairly accurate frame rate. In second phase a rotation, translation, scaling and orientation invariant feature extraction technique has been establish to take out the feature of the input image based on moment feature extraction method. Gesture (Gesture) means a association of element of body. Gesture Recognition is the technique that recognizes the movement of any part of body. It recognizes hand, head, arms or some part of the body. Hence the objective of Gesture Recognition is to give interface of human body through computer via mathematical algorithm. This paper provides a real time vision based technique for recognizing human body action using MATLAB. It in addition provides the working features of recognizing procedure utilize Skin detection algorithms and Edge detection.