Virtual Rehabilitation Training System Based on Surface EMG Feature Extraction and Analysis

Aiming at the characteristics that electromyography (EMG) signals can reflect the human body’s motive intention and the information of muscle’s motive state, this paper makes a thorough study on the evaluation of surface electromyography signals’ motive state. At the same time, EMG signals can reflect the characteristics of limb movement and its changing rules, and can acquire the functional characteristics of limb movement so as to accurately evaluate the rehabilitation status of patients. In this paper, EMG signal analysis and feedback control are introduced into the virtual rehabilitation system to study the methods of EMG parameter identification and dynamic feature extraction, and obtain the EMG characteristics and variation rules related to human motion patterns. In this paper, a rehabilitation training system based on EMG feedback and virtual reality is built, and the validity of the system is verified by patient experiment. The feasibility of the system is verified by the methods of validity of the algorithm, recognition rate of the system action pattern and fatigue evaluation.