Human Arm Movements Recognition Based on EMG Signal

Abstract

The EMG signal which is generated by the muscle's activity, diffuses to the skin surface of human body. This paper describes the EMG signal, specification and characteristic of this signal. Moreover, analysis of the processed signal is done to select the suitable features (amplitude and period) that can be used in the neuroidentifier. A proposed structure is used for macrocalcification and neruo-recognition for human arm muscles-based EMG signal. EMG simulator is presented with suitable simple modifications. This simulator is used to generate an EMG data related to clinical specification, and hence these data are used to simulate an EMG signal using EMGlab software. The proposed structure is simulated using MATLAB Ver. R2009a, and satisfied results are obtained, that explains the ability of the proposed structure to classify and recognize the movements of human arm-based EMG signals.