As a typical biomedical signal, the electromyography (EMG) is now widely used as a human-machine interface in the control of robotic rehabilitation devices such as prosthetic hands and legs. Immediately detecting and eliciting of a valid EMG signal are greatly anticipated for ensuring a fast-response and high-precision EMG control scheme. This paper utilizes two schemes, Teager-Kaise Engergy (TKE) operator and Morphological Close Operation (MCO), to improve the accuracy of the onset/offset detection of EMG activities. The TKE operator is used to amplify the EMG signal’s amplitude change on the initiation/cessation phases, while the MCO is adopted to filter out the false positives of the binary sequence obtained by the fore TKE operation. This method is simple and easily to be implemented. After selecting appropriate filtering parameters (T1, T2 and j), it can achieve precise onset detection (absolute error <10ms) over a variety of signal-to-noise ratios (SNR) of the biomedical signal.
CITATION STYLE
Yang, D., Huang, Q., Yang, W., & Liu, H. (2015). EMG onset detection based on Teager–Kaiser energy operator and morphological close operation. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 9244, pp. 257–268). Springer Verlag. https://doi.org/10.1007/978-3-319-22879-2_24
Mendeley helps you to discover research relevant for your work.