Hand Gesture Recognition Using Artificial Neural Network

Abstract

Gesture is one of the most easy and expressive ways of communications between human and computer in a real world. In our day to day life we use various gestures to represent our intention. Hand and face gesture are most important methods for nonverbal communication for human beings. Development in hand gesture recognition using machine interface has got great improvements in recent years. A system is developed for hand gesture recognition in mat lab by using an artificial neural network. Here a dataset is collected with various gestures and through color based segmentation and various morphological operations gestures are segmented from input image. Here feature extraction takes place on the basis of Histogram of Gradient method. After segmentation classifies the gestures into left hand and right hand gestures along with labels. This will passes to the neural network for training and train-rp is the function used for training the gestures. And during the testing face video is given to the system and the system will recognize the corresponding output in the frame and predict it.