T. Grujić and V. Zanchi (Croatia)
Electromyogram, Wavelets, Denoising, MuscleActivation Intervals, Threshold
In this work two problems concerning surface myoelectric signals were studied. First of them is the rejection of motion artefacts and white noise from emg signals recorded during gait. Several filtering techniques were tested and compared: band pass filtering procedures and wavelet based methods which use discrete wavelet transformation and application of thresholds. Several different wavelets and threshold methods will be employed and results will be compared. The other scope of this work is the detection of muscle activation intervals from emg signals recorded during gait. The threshold method for measurement of active intervals will be proposed. The method normalises emg's linear envelopes according to 100% of maximum voluntary contraction value (MVC) and then employs the threshold to detect muscle activity. It will be shown that if the method was applied on emg signals denoised by wavelet procedure, it provided satisfying results.
Important Links:
Go Back
