Recently, there have been many studies to bionic leg using physical sensor. But existing bionic leg have to move the same speed as training stage regardless of person’s intention. To solve this problem, a few sEMG signal based gait phase recognition...
Recently, there have been many studies to bionic leg using physical sensor. But existing bionic leg have to move the same speed as training stage regardless of person’s intention. To solve this problem, a few sEMG signal based gait phase recognition studies are presented. However they didn’t supply real-time recognition for gait phase. In this paper, we propose a gait phase prediction using linear interpolation and user adaptive classification based on sEMG signal. Experimental results show that the average accuracy of user adaptive classification is about 81.4% whereas that of existing method is about 71.22%.