IIUM Repository

Design and performance analysis of artificial neural network for hand motion detection from EMG signals

Ibrahimy, Muhammad Ibn and Ahsan, Md. Rezwanul and Khalifa, Othman Omran (2013) Design and performance analysis of artificial neural network for hand motion detection from EMG signals. World Applied Sciences Journal , 23 (6). pp. 751-758. ISSN 1818-4952

[img]
Preview
PDF - Published Version
Download (620kB) | Preview

Abstract

Besides prosthetic device control and neuromuscular disease identification, electromyography (EMG) signals can also be applied in the field of human computer interaction (HCI) system. This article represents the classification of Electromygraphy (EMG) signal for the detection of different predefined hand motions (left, right, up and down) using artificial neural network (ANN). The neural network is of backpropagation type, trained by Levenberg-Marquardt training algorithm. Before the classification process, the EMG signals have been pre-processed for extracting some features from them. The conventional and most effective time and timefrequency based features are extracted and normalized. The neural network has been trained with the normalized feature set with supervised learning method. The obtained results show that the designed network can successfully classify the hand motions from the EMG signals with the success rate of 88.4%. The performance of the designed network has also been compared to similar research work, whereby it certainly shows the outperformance.

Item Type: Article (Journal)
Additional Information: 4637/30592
Uncontrolled Keywords: Electromyography; Wavelet Transform; Artificial Neural Network; Backpropagation; Classification; Human Computer Interaction
Subjects: T Technology > T Technology (General)
Kulliyyahs/Centres/Divisions/Institutes (Can select more than one option. Press CONTROL button): Kulliyyah of Engineering
Depositing User: Dr Muhammad Ibrahimy
Date Deposited: 11 Jul 2013 11:51
Last Modified: 11 Jul 2013 11:51
URI: http://irep.iium.edu.my/id/eprint/30592

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year