Keywords: Myo armband, finger movement detection, surface electromyograms, virtual piano
We analyse the feasibility of finger activity detection by the Myo armband. The latter is placed on the forearm of the healthy male subject and measures eight surface electromyogram (EMG) channels from different finger extensors and flexors, except the thumb. We systematically analyse different finger movements. Our algorithm detects the extensions of fingers with accuracy of 95 %, whereas detection of other finger movements proves to be less accurate. Movements of two fingers decrease the accuracy of movement detection to 50 %. Quality of finger detection depends on the flexibility of subject’s fingers, electrical resistance between the skin and the EMG sensors and placement of the Myo armband. We demonstrate the usage of finger activity detection with a virtual piano application.
Open this page, put Myo on left forearm and connect armband with computer. (Note: this parameters work good for me, you must change some parameters).
In the future: Create more effective way for detection fingers. (Maybe support vector machine).