퓨전 센서를 이용한 EMG 신호 기반 실시간 보행 패턴 인식 시스템의 설계

초록

In this paper, we design an EMG signal based real-time locomotion mode pattern recognition system using fusion sensor. Detection timing for recognizing locomotion mode are determined by fusion sensor information. Feature extraction of EMG signal and classification process are performed in real-time. The development of the system requires hardware and software co-design. Since the frequency of EMG signals ranges between 1 and 500Hz, we design the wireless measurement device that can perform at least 1,000Hz sampling for analyzing EMG signal. Real-time pattern recognition software based on desktop PC is modularly designed for applying various feature extraction methods and classification methods. This design of system can be utilized in developing medical equipment for rehabiliation.