连续肌肉电信号的辨识

RECOGNITION OF CONTINUOUS EMG

以前对肌电信号的处理都是针对手工分割后的信号进行辨识，但实际应用中需要处理的是连续肌电信号。本文中采用了时域参数与频域参数相结合的方法对连续的肌电信号进行自动分割，然后再辨识分割后的肌电信号，取得了很好的结果。

In the past, only hand divided EMG can be recognized, while in practical cases, EMG recorded from the surface electrodes is continuous. Parameters of both time domain and frequency domain are introduceds to divided the continuous EMG into a series of segmented EMG and then recognize them. From experimental results, the recognition ratio is high and the method proposed in this paper is very effective.