Classification of Hand Gestures with Real Time Muscle Activity Detection for Myoelectric Control of Upper Limb Prosthesis

Abstract

This paper presents the classification of basic hand movements with the determination of onset and offset timings of muscle activity in real time using surface Electromyography (sEMG). Integration Profile (IP) method is evaluated to detect muscle activity in real time. Dynamic Time Warping (DTW) is used to classify hand movements using detected muscle activity signals. The sEMG data collection, muscle activity detection and classification of different muscle activities are performed in the National Instrument (NI) LabView environment. The movement classification results suggest that the proposed procedure accurately demonstrates the importance of muscle activity detection in the classification. © 2023 IEEE.