Real-time control signal extraction based on instantaneous power of surface electromyogram

The previous studies on Human-machine Interface (HMI) based on surface electromyogram (SEMG) control seldom considered individual difference. In this paper, we proposed a novel method to distinguish the handedness of subjects, and extract the real-time control signals from SEMG. SEMG signals recorded at left calf, right calf, left shoulder and right shoulder were analyzed. The instantaneous power of SEMG signals were calculated based on Filter technique and short-time Fourier transformation (STFT).The handedness was discriminated to determine the proper threshold for different subjects. The real-time control signals were generated automatically by comparing the instantaneous power and the thresholds preset.