Decoding of hand shapes based on electromyographic signals during playing guitar chords

Abstract

In this paper, aiming to study on a new entertainment direction, we investigated the decoding performance of complicated hand shapes from ElectroMyoGraphic (EMG) signals during playing guitar chords. Three healthy right-handed subjects participated in this experiment. During the experiment, they played four guitar chords (The major chords of 'C', 'F', 'G', and 'A') as well as being relaxed. The EMG signals were recorded from the left forearm by using 12 surface EMG electrodes. By using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), the decoding performances could be more than 95% for all subjects. Our results will encourage the development of surface EMG based entertainment systems. © 2011 Springer-Verlag.