Hand gestures recognition from multi-channel forearm EMG signals

Abstract

The control system based on the surface Electromyography (sEMG) signal provides a wireless, convenient and natural choice to Human Computer Interaction (HCI). The identification of human hand gestures can offer enough kinds of controlling commands to intelligent devices in real time. In order to improve the classification accuracy in recognizing hand gestures, this paper explored the signal acquisition, signal processing and feature extraction methods of 6-channel forearm EMG signals. By utilizing Chebyshev II filter (25– 450 Hz), 9 time domain features in sliding windows, PCA algorithm and SVM classifier, 17 hand gestures (HG), which include 6 wrist actions (WR) and 11 finger gestures (FG), are recognized with the accuracy of more than 95%.