A technique for classification and decomposition of muscle signal for control of myoelectric prostheses based on wavelet statistical classifier

Since surface electromyography is an electrical activity of superficial muscles and is an essential tool to investigate assessments protocols to be required for prosthetic design, so here, the wavelet transforms based interpretation of Surface Electromyogram signal for classifications of upper arm operations were investigated. The study presented methods of processing and analyzing Surface Electromyogram signal for upper arm motions for extracting accurate patterns of the signal. From these recorded signals, amplitude estimated features were extracted and explored significantly. Then a comparative study to evaluate the wavelet denoising for optimal motor unit action potential detection through the decomposition based on the different wavelet functions of Daubechies, Coiflet and Symmlets families were investigated and tabulated. Thereafter linear discriminating analysis pattern classifier approach was employed to analyze classification performance for different upper arm movements. Results inferred that Daubechies wavelet families were more suitable for the analysis of surface electromyogram signals of different upper arm motions and a classification accuracy of 85.0% was achieved. Finally data projection method of analysis of variance technique was implemented for the effectiveness of recorded surface electromyogram signals for class separability of upper arm motions.