Gesture Recognition System

Abstract

Gesture based interaction systems are becoming more and more prominent both at work environment and home. The projects expects to build up a system that can perceive hand gesture which can be utilized as a input to associate with the PC which can be connected to various operation of PC. One of the key zones that should be taken a gander at while developing such systems is the image processing stage.. We have proposed a technique for real Hand Gesture Recognition and feature extraction utilizing a web camera. In this approach, the picture is gotten through digital camera related to associated with the system. To start with ,the input picture is preprocessed and threshold value is used to remove noise from it and smoothen the image. After this apply region filling to fill holes in the object of interest. This helps in improving of the classification and recognition step. At that point expel all little object, this is done to evacuate additional undesirable noise or object from picture. At the point when the preprocessing is finished the image is passed on to feature extraction stage. For feature extraction “HU moments” are used because of their distinct properties like scale rotation and translation invariance. The exaggerated elements are normalized and composed to tell the comparing gesture. The classification results are displayed to user. The purpose of this is to improve the accuracy of classification.