Preliminary Study on Finger Gestures for Surface Electromayograph (sEMG) based Number Recognition

Wide use of surface electromyograph (sEMG) has been made for efficient recognition of finger gestures due to its convenient to use and distinguishing signal patterns along finger movements. For high classification accuracy, it is important to have a consistent feature of sEMG signal feature for each finger gesture of a number. However, feature of sEMG signal for identical finger gesture can be different because of different muscle activation even in identical shape of gesture. Thus the experimental results in this work may motivate further study on how to make consistent muscle activation for a finger gesture at each try.