Surface Electromyogram signal estimation based on wavelet thresholding technique

摘要: Surface Electromyogram signal collected from the surface of skin is a biopotential that may be influenced by different types noise. This considerable drawback in processing sEMG signals. To acquire clean contains useful information, we need to detect and eliminate these unwanted parts signal. In this work, new method based on wavelet thresholding technique presented which provides an acceptable purified signals for study are extracted various hand movements. We use three movements calculate near optimal estimation parameters. work two techniques, namely Stein unbiased risk (SURE) estimator adaptive Bayes utilized coupled with selected mother wavelets levels decomposition. After designing technique, evaluating efficacy method, formed sent pattern recognition system order discriminate among eight The acquired results indicate using SURE approach appropriate producing without noise result improvement application movement recognition.