Lower Limb Electromyography Signal Denoising Via EMD and Improvedwavelet Threshold

The surface electromyography (sEMG) signal can reflect the human neuromuscular motion information. The study of lower limb electromyography signal provides a scientific and effective basis for the rehabilitation evaluation of lower limb injury. Pure sEMG is an important prerequisite for evaluating the rehabilitation status of lower limbs. In order to effectively reduce the noise mixed in sEMG signals, this paper proposes a de-noising method based on empirical mode decomposition combined with improved wavelet threshold. Firstly, the lower limb EMG signal is decomposed by empirical mode decomposition to obtain the IMF component. The high-frequency component is analyzed by improved wavelet threshold analysis, and the low-frequency component is not processed. Finally, the processed high-frequency component and low-frequency component are reconstructed to obtain the denoised signal. The experimental results show that the method proposed in this paper can retain effective information, has high signal-to-noise ratio and good practicability and denoising effect.