Onset determination of muscle contraction in surface electromyography signals analysis

This paper introduces the online and offline onset detection problem of muscle contraction and their specialties, and gives some criteria of the performance of onset detection algorithms, then several methods or algorithms in commonly used are analyzed and compared, such as visual decision, mean value comparison, linear regression of low pass filter and sliding window average. And it discusses the decision threshold issue and presents a new maximum value detection algorithm (MVD) based on the bipolar model algorithm is fast and insensitive to background noise, and fit for online detection.