Abstract

Physiological dielectric characteristic signal such as electromyography (EMG) signal or electrocardiogram (ECG) signal is the basic data of human physiology in researching and monitoring athletics training or recovery exercise. After the wave decomposition to biological signal, in order to extract the feature of signal and eliminate the noise in the superior wave, based on wavelet analysis, using global statistics and local maximum likelihood analysis to the wavelet series respectively, a method to filter the signal was given here, thus the recognition and extraction of human EMG-signal mixed with random noise and ECG-signal was achieved, at last, relevance analysis was made to the results, which test the feasibility of this method of recognition and extraction. Verf.-Referat