No abstract

The study improve sEMG signals, especially, for the transradial amputees, it investigated the classification accuracy of surface Electromyography (sEMG) with low-cost Myo armband device. Seventeen specific viable gestures used in daily-life were compared between two groups, the first one was represented by six transradial amputees and the second one was represented by six able-bodied subjects. In addition, the measured signals from forearm muscles were utilized to drive 3D printed prosthetic hand controlled by Pattern Recognition (PR) technique. In order to analyze the acquired sEMG signals, the PR system employed which consists of three main parts: segmentation, feature extraction and classification. The LDA classifier was employed in order to obtain optimum accuracy of the system. The classification accuracy of Myo armband showed high performance where the achieved classification accuracy for abled bodies subjects was more than (94.85%) with LDA classifier and for transradial amputees subjects (88.2%) with LDA classifier. Also, it was acceptable for 3D printed hand controller for 7 movements in study case of transradial amputees. The results may conclude that low cost Myo armband can be applicable for controlling 3D printed prosthetic hand to distinguish hand movements.