EMG-based Human Machine Interface Control

Abstract

The use of EMG signals has been gaining prevalence in prosthetic control and gesture-control applications. This study aims to develop an EMG-based human machine interface system. A Myo Armband with eight electrode pairs is worn by a user to acquire and wirelessly transmit EMG data to a central controller. A pattern recognition algorithm is implemented on a central controller to recognize three different hand gesture commands. As a demonstration, a camera system equipped with servo motors is built. The recognized commands can remotely pan each camera and select one of multiple video feeds to display. The study demonstrates that the EMG-based pattern recognition could be a viable human machine interface option for a broad range of applications in industrial, medical, and consumer markets.