PERFORMANCE ENHANCEMENT OF THE SURFACE ELECTROMYOGRAPHY SIGNAL USING HYBRID FEATURES SELECTION WITH AN APPLICATION ON MOVING ROBOTIC MANIPULATOR

Authors

  • Sadiq J. Abou-loukh Department of electrical engineering, College of Engineering, University of Baghdad, Baghdad 10001, Iraq.
  • Ibraheem Kasim Ibraheem Department of Computer Engineering Techniques, Al-Mustaqbal University College, 51001 Hilla, Iraq.

Keywords:

Myo Gesture Armband, Electromyography Signal, Robotic Arm, Feature Extraction, DUE Arduino, Pattern Recognition

Abstract

The objective of this paper is to find a hybrid feature set that obtains good descriptive information on surface Electromyography signals to move a 4-DoF robotic hand. The hybrid feature set is obtained from the investigation of the performance of the Time Dependent Power Spectrum Descriptors(TD-PSD) feature set which consists of ( , , , , , , and TD feature set. The proposed system was tested on datasets extracted from six healthy subjects and the results showed that the support vector machines achieved higher system accuracy with 95.61 % but with a high processing time of 2.48 sec compared to Linear Discriminant Analysis with an accuracy of 93.29% and low time processing of 0.573sec and 93.81% accuracy was achieved by K-Nearest Neighbor (K-NN)with a processing time of 1.332sec for hybrid feature set which consists of Mean Absolute Value, Root Mean Square, log(m0), and Waveform Length.

Downloads

Download data is not yet available.

References

a

Downloads

Published

2023-11-19

How to Cite

Sadiq J. Abou-loukh, & Ibraheem Kasim Ibraheem. (2023). PERFORMANCE ENHANCEMENT OF THE SURFACE ELECTROMYOGRAPHY SIGNAL USING HYBRID FEATURES SELECTION WITH AN APPLICATION ON MOVING ROBOTIC MANIPULATOR. Operational Research in Engineering Sciences: Theory and Applications, 6(3). Retrieved from https://oresta.org/menu-script/index.php/oresta/article/view/626

Issue

Vol. 6 No. 3 (2023): Operational Research in Engineering Sciences: Theory and Applications

Section

Articles