A Robust sEMG Feature Selection based on Recursive Feature Elimination for Gait Recognition

초록

This paper aims to propose a novel approach for gait phase recognition using optimal feature subset based on recursive feature elimination. This study has collected the large number of gait data to improve the reliability of quantitative assessment of natural variability associated with muscle activity during free walking. The gait system was designed using a 9-channel electromyography sensor for measuring the data. Gait recognition method was proposed using support vector machine based on recursive feature elimination to determine best feature subset. The preliminary experimental results show that, the proposed method achieves the robust classification performance.