Publication Type : Journal Article
Source : 2nd International Conference on Future Learning Aspects of Mechanical Engineering (FLAME - 2020), Amity University, Noida, 2020
Url : https://link.springer.com/chapter/10.1007/978-981-33-4320-7_78
Campus : Coimbatore
School : School of Engineering
Department : Mechanical Engineering
Year : 2020
Abstract : Electromyography (EMG) signals have been used in clinical diagnosis and rehabilitation owing to the information that the signal carries about motion intent. However, EMG signals are inherent to noise which degrades the performance of classifiers. In the present study, denoising of the EMG signal was studied with its effect on the classification accuracy of sit to stand (STS) phases. Four different phases of STS task were marked with the help of knee and trunk angular deviation data as acquired during the experimentation on five healthy participants. Two different denoising methods were considered; one method (D1) deals with seventh-order Daubechies wavelet denoising, while in the second method (D2), Teager–Kaiser energy operator (TKEO) was applied over the previously denoised EMG signal in D1. Method D2 was found to improve the overall accuracy of the K-nearest neighbors (KNN) classifier with the highest improvement in the true positive rate of intention phase (Phase II) of STS task.
Cite this Research Publication : S. Bhardwaj, A.A. Khan, and M. Muzammil, “Effect of EMG denoising on classification accuracy of sit to stand phases,” in 2nd International Conference on Future Learning Aspects of Mechanical Engineering (FLAME - 2020), Amity University, Noida, 2020.