Sensor Based Hand Gesture Identification for Human Machine Interface

This paper presents innovative and simple methods for extracting essential hand motion segmentation features from electrode signals obtained from the abductor longus muscle and human-machine interface (HMI) electromyography (EMG) signals in individuals with spinal cord injuries (SCI). The filtering process removed unwanted thumb-over-elbow sound, and distinct feature sets were derived. These feature sets are employed to control both the wheelchair’s movements and an automated tool. Lastly, the study demonstrates that using SCI SEMG signals in an HMI, an individual can effectively control the cursor and navigate the wheelchair in the desired direction, resulting in a more user-friendly interface.