EMG Signal-Based Gait Phase Detection Using a EMG Signal Graph Matching(ESGM) Algorithm

초록

This paper presents a gait sub phase detection method using electromyogram (EMG) with a EMG signal graph matching (ESGM) algorithm. Existing pattern recognition and machine learning using EMG signal has several innate problems in trying to control a lowerJimb power-assist robot. In general, the gait sub phase is recognized as at least four gait iteps. With respect to time domain features, their feature values may be analogous because fwo different gait steps may have similar muscle activation, Therefore, it is not easy to select proper feature set of EMG signals. The ESGM algorithm reflects better timing characteristics of EMG signals than the time domain feature extraction algorithm. Therefore, it can provide high accuracy of gait phase detection. The experimental results show that the average true positive accuracy of the proposed method is2loh better than that of the