Pattern Recognition in Surface Electromyography

Abstract

In this paper, a method is proposed to differentiate patterns in one-channel surface electromyography. This particular method uses an example signal containing all relevant patterns to learn spectral coefficients. These coefficients are used to analyze unknown signals. An example of finger recognition algorithm based on a signal from forearm muscles is shown and discussed.