R-CNN and wavelet feature extraction for hand gesture recognition with EMG signals
- PDF / 1,804,821 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 67 Downloads / 196 Views
(0123456789().,-volV)(0123456789().,-volV)
ORIGINAL ARTICLE
R-CNN and wavelet feature extraction for hand gesture recognition with EMG signals Vimal Shanmuganathan1 • Harold Robinson Yesudhas2 • Mohammad S. Khan3 Amir H. Gandomi5
•
Manju Khari4
•
Received: 25 March 2020 / Accepted: 7 September 2020 Springer-Verlag London Ltd., part of Springer Nature 2020
Abstract This paper demonstrates the implementation of R-CNN in terms of electromyography-related signals to recognize hand gestures. The signal acquisition is implemented using electrodes situated on the forearm, and the biomedical signals are generated to perform the signals preprocessing using wavelet packet transform to perform the feature extraction. The R-CNN methodology is used to map the specific features that are acquired from the wavelet power spectrum to validate and train how the architecture is framed. Additionally, the real-time test is completed to reach the accuracy of 96.48% compared to the related methods. This kind of result proves that the proposed work has the highest amount of accuracy in recognizing the gestures. Keywords R-CNN EMG signal Wavelet power spectrum Discrete wavelet transform Gesture recognition Validation
1 Introduction & Mohammad S. Khan [email protected] Vimal Shanmuganathan [email protected] Harold Robinson Yesudhas [email protected] Manju Khari [email protected] Amir H. Gandomi [email protected] 1
Department of Information Technology, National Engineering College, Kovilpatti, India
2
School of Information Technology and Engineering, Vellore Institute of Technology, Vellore, India
3
Department of Computing, East Tennessee State University, Johnson City, USA
4
Department of CSE, Ambedkar Institute of Advanced Communication Technologies and Research, New Delhi, India
5
Faculty of Engineering and Information Technology, University of Technology Sydney, Sydney, NSW 2007, Australia
Human–machine interfaces (HMI) [1] might potentially expand the quality of life of disabled persons who suffer from neuromuscular diseases. The interactions between a machine and a human will need a spontaneous, healthy and high information/transmission ratio for these kinds of defects [2]. An HMI framework normally performs the feedback- and control-based information exchange to implement the 2-directional communication between a machine and a human. Many bio-based signals are utilized, like MMG [3], ultrasound [4] and EEG [5], to implement this kind of control. Within these kinds of bio-related signals, the electromyography (EMG) signal is mostly used for HMI. EMG-related systems [6] are commonly utilized in different applications. Moreover, the control mechanisms are constructed using simplified methods. The EMG signals [7] are stored in the antagonistic muscles for every channel to construct a degree of freedom. The signal is picked up as the amplified at the electrode field that the signal could be utilized to diminish the frequency noise; it is normally rectified into a standard format to identify the EM
Data Loading...
