EMG Signal Filtering Based on Variational Mode Decomposition and Sub-Band Thresholding

Surface electromyography (EMG) signals are inevitably contaminated by various noise components, including powerline interference (PLI), baseline wandering (BW), and white Gaussian noise (WGN). These noises directly degrade the efficiency of EMG processing and affect the accuracy and robustness of further applications. Currently, most of the EMG filters only target one category of noise. Here, we propose a novel filter to remove all three types...