Abstract
This work is focused on segmentation of electromyographic (EMG) signals. The segmentation of the EMG signal is necessary for further processing and creation of muscle activations patterns.
Processed signals were measured during periodic movements: cross-country skiing and walking. Segmentations based on the linear EMG envelope and on the video recording are described thoroughly and compared.
The detected time intervals of movement cycles calculated by using both methods are compared. The results are evaluated by comparison of the muscle activity intervals.
Our conclusion is that both methods provide comparable results. Video processing is more difficult but provides segmentation independent of the EMG signal.
Thus, video-based segmentation is favorable especially if the quality of the recording allows for easy derivation of the segmentation signal.