SEMG based classification of hand gestures using artificial neural network

Abstract A Human Machine Interface (HMI) is an application that permits the correspondence between a man and a machine. Surface Electromyogram (sEMG) is an electrophysiological sign that gives inalienable data about the exercises of the human skeletal muscle. This paper proposes an sEMG design acknowledgment framework to control the myoelectric hand framework utilizing sEMG for HMI with the assistance of neural systems. Six parametric element extraction calculations are applied to get the unmistakable data from sEMG, for example, AR (Autoregressive) Burg, AR Yule-Walker, AR Covariance, AR Modified Covariance, Levinson Durbin Recursion, and Linear Prediction Coefficient. The sEMG signals are demonstrated utilizing General Regression Neural Network (GRNN), Probabilistic Neural Network (PNN) and Radial Basis Function Neural Network (RBFNN). The reaction of the HMI framework has gotten normal mean arrangement exactness of 94.04% for AR Burg highlights utilizing RBFNN. From the outcomes, it is seen that AR Burg utilizing the Radial premise neural system outflanked different systems in arranging twelve distinctive finger developments and females control the framework simpler than guys. It is likewise apparent that the age bunch between 26 and 30 delivered better exhibitions.