sEMG signal detector using discrete wavelet transform

This article presents a sEMG signal detector by means of different applications of the Discrete Wavelet Transform. A training system and protocol is used to acquire the sEMG signals. The last of four training stages is used. These signals are processed, segmented and labeled, automatically, as three different levels of contraction. The level's order (1-2-3) varies from one session to another. The automatic segmentation and label proved to work with the 20 sessions. Segmentation and labeling comparison of 4 training sessions' results is shown. The used algorithms provided good results with all 20 sessions.