근전도 기반 보행단계의 특징추출 및 유각기 분류 기법

초록

In this paper, we propose a swing phase classification and its feature extraction method to classify flexion and extension of tested legs in gait phase for level walking. The effect of using six muscles in femoral muscle, calf muscle pots and glutaeus maximus pots and 10 features(RMS, VAR, MAV, MAV1, SSC, SSI, ZC, IEMG, WAMP, WL) are evaluated. Experimental results show that the average classification accuracy of LDA(Linear Discriminant Analysis) is 84.2% when vstus lateralis muscle and SSC(Slope Sign Change) are used.