Abstract

Comparative study of myoelectric pattern recognition using SVM and PNN classifiers based on wavelet analysis

Author(s): Firas AlOmari, Guohai Liu

The choice of a proper wavelet familywith a fast and robust classifier is an important step in the construction of a myoelectric control pattern recognition system for a prosthetic hand. In this study, five hand motions were classified by using sixwavelet functions extracted features fromsEMG signals. The selected wavelet families that were used to decompose the recorded sEMG signals are Biorthogonal (bior). Coiflet (coif), Daubechies (db), and Symmlet (sym). Two different recognitionmethodswere employed for classification procedure: support vector machine (SVM), probabilistic regression neural network (PNN). The results of our experiment demonstrate that the use of wavelet families at a high decomposition level increases the recognition rate of hand motions. The highest achieved classification rate was 96%, by using the PNN classifier based on coif4 at the sixth decomposition level.