Improved DTCWT-LMS and FastICA Based sEMG Signals Filtering
Lin Li, Wang Jianhui, Fang Xiaoke, Shusheng Gu · International Journal of Signal Processing Image Processing and Pattern Recognition · 2013
A novel design of dual-tree complex wavelet transform (DTCWT) and fastICA was proposed, aiming at the noise interference and aliasing between multi-channels sEMG signals. Firstly, DTCWT was utilized to decompose signals to different frequency band. Secondly, an improved LMS adaptive filter was designed for filtering sub band noise layer by layer. Finally, fastICA algorithm was introduced to separate crosstalk between channels. Some experiments were carried out to compare the proposed method with other algorithms, and the results showed that the algorithm proposed could filter noise effectively, keep better convergence especially in low signal-to-noise ratio and eliminate crosstalk more thoroughly by fastICA.