Study and analysis of surface EMG for the lower limb prosthesis

Wavelet transformation (WT) and wavelet packet transformation (WPT) are used in this paper to eliminate the surface EMG noises sampled from the lower limb during the subjects of study walked normally on the flat by comparing the denoised results, WPT wins an advantage of WP with the same threshold. And the result shows that if the suitable scale and the threshold are selected, the high-frequency noises can be eliminated effectively. According to the characteristic of WPT, the motion features are extracted during the stance phase and the swing phase by using the WPT. The figures show the fine extraction result in this paper. Lastly, the LVQ network is used to classify the motion features.