Abstract
This paper proposes a detection method of tongue motions based on EMG signal around a frontal neck, which is used to control a support device for quadriplegic patient. This method extracts some features of EMG signal that is measured by multiple array electrodes attached on a skin surface of a frontal neck. A Support Vector Machine (SVM) in the method classifies the signal into six kinds of the tongue motion, since the SVM is a potent pattern recognition algorithm in weakly nonlinear system. We investigate the performance of the detection algorithm through the experiments and then confirmed that accuracy of all motions' detection is more than 92%.
