Surface EMG Based Hand Manipulation Identification Via Nonlinear Feature Extraction and Classification

This paper proposes and evaluates methods of nonlinear feature extraction and nonlinear classification to identify different hand manipulations based on surface electromyography (sEMG) signals. The nonlinear measures are achieved based on the recurrence plot to represent dynamical characteristics of sEMG during hand movements. Fuzzy Gaussian Mixture Models (FGMMs) are proposed and employed as a nonlinear classifier to recognise different hand grasps and in-hand manipulations captured from different subjects. Various experiments are conducted to evaluate their performance by comparing 14 individual features, 19 multifeatures and 4 different classifiers. The experimental results demonstrate the proposed nonlinear measures provide important supplemental information and they are essential to the good performance in multifeatures. It is also shown that FGMMs outperform commonly used approaches including Linear Discriminant Analysis, Gaussian Mixture Models and Support Vector Machine in terms of the recognition rate. The best performance with the recognition rate of 96.7% is achieved by using FGMMs with the multifeature combining Willison Amplitude and Determinism.