Research on Improved Wavelet Denoising Method for sEMG Signal

When using the upper limb sEMG signal to recognize the upper limb motion, a complete and effective sEMG signal denoising method is needed to improve the recognition rate. The wavelet denoising method is commonly used at present. Through experiments, an improved sEMG signal denoising method based on wavelet analysis is proposed. This method that combines the traditional wavelet threshold denoising (WTD) method with the wavelet digital filter threshold denoising (WDFTD) method makes up the deficiency of the traditional wavelet denoising method. To meet the needs of the selection of an optimal wavelet function, decomposition layer number and threshold function, a comprehensive index P is proposed to evaluate the quality of denoising. Finally, an improved wavelet denoising (IWD) method based on wavelet analysis was obtained. Compared with traditional wavelet denoising method, the P index of this method was 0.2341, which proves that its denoising effect was better than that of the conventional wavelet denoising method.