Human hand gesture detection based on EMG signal using ANN

There has been wide range of expansion in hand gestures close in style, pattern, movement and feature as well as in meaning. From that aspect this following paper confabs an improved version of hand gesture recognition based on segmentation process and backpropagation algorithm through the analysis of myoelectric signals generates from the movement of brachioradialis muscle and antebrachial vein of lower portion of forearm as well as flexor carpi radialis tendon and flexor digitorum superficial tendons of upper portion of wrist connected to forearm. From the received forearm electromyography (EMG) signals separation of individual signals through segmentation and corresponding evaluated values are used as the input of feed forward artificial neural network which would act as a classifier to obtain required distinguishable signals or commands to communicate. The neural network is trained with number of signals of different people and tested with other signals for minimum errors as well. Without considering depth or distance this proposed method will provide a good variety of classified signals for small changes in muscular signal changes for patient monitoring through virtual command.