Hand Gesture recognition is technology which interpret human gestures using various algorithms.Interpreting human hand gestures has various challenges and issues such as image noise,visibility and orientation. There are various kinds of computer based algorithms have been proposed in the literature to overcome these limitations and still needs improvement. Hence in this research work a new Hand Gesture recognition system for augmented reality using Genetic Algorithm and Artificial neural network algorithm has been proposed.This shows that the experiment is successful and gesture recognition system is vigorous against various changes that are made in illumination changes and background changes. Experimental results show that the extracted features are effective, robust, and can cover the entire feature space of the selected gestures. This method satisfactory performance when compared with convensional methods