TY - JOUR
TI - Implementation of a real-time automatic onset time detection for surface electromyography measurement systems using NI myRIO
AU - Lersviriyanantakul Chaiwat
AU - Booranawong Apidet
AU - Sengchuai Kiattisak
AU - Phukpattaranont Pornchai
AU - Wongkittisuksa Booncharoen
AU - Jindapetch Nattha
JN - Thermal Science
PY - 2016
VL - 20
IS - 12
SP - 591
EP - 602
PT - Article
AB - For using surface electromyography (sEMG) in various applications, the  process consists of three parts: an onset time detection for detecting the  first point of movement signals, a feature extraction for extracting the  signal attribution, and a feature classification for classifying the sEMG  signals. The first and the most significant part that influences the  accuracy of other parts is the onset time detection, particularly for  automatic systems. In this paper, an automatic and simple algorithm for the  real-time onset time detection is presented. There are two main processes in  the proposed algorithm; a smoothing process for reducing the noise of the  measured sEMG signals and an automatic threshold calculation process for  determining the onset time. The results from the algorithm analysis  demonstrate the performance of the proposed algorithm to detect the sEMG  onset time in various smoothing-threshold equations. Our findings reveal  that using a simple square integral (SSI) as the smoothing-threshold  equation with the given sEMG signals gives the best performance for the  onset time detection. Additionally, our proposed algorithm is also  implemented on a real hardware platform, namely NI myRIO. Using the  real-time simulated sEMG data, the experimental results guarantee that the  proposed algorithm can properly detect the onset time in the real-time  manner.