Estimation of finger postures to control a maniform device for playing a trumpet using electromyographic signals with external triggers

Electromyographic (EMG) signals have been used to control active prosthetic arms for amputees. One of the obstacles in making such prosthetic arms is the timed estimation of posture, because EMG signals and muscle movements are not necessarily synchronized. We estimated the finger motions for trumpet players by using both surface EMG (sEMG) and the timing information using body motion. The algorithms consisted of Principal Component Analysis (PCA), and Support Vector Machine (SVM). The results showed that applying the timing information using body motion increases how precisely the motion of the fingers is estimated.