0

Full Content is available to subscribers

Subscribe/Learn More  >

A Performance Comparison of EMG Classification Methods for Hand and Finger Motion

[+] Author Affiliations
Sungtae Shin, Reza Langari

Texas A&M University, College Station, TX

Reza Tafreshi

Texas A&M University at Qatar, Doha, Qatar

Paper No. DSCC2014-5993, pp. V002T16A008; 7 pages
doi:10.1115/DSCC2014-5993
From:
  • ASME 2014 Dynamic Systems and Control Conference
  • Volume 2: Dynamic Modeling and Diagnostics in Biomedical Systems; Dynamics and Control of Wind Energy Systems; Vehicle Energy Management Optimization; Energy Storage, Optimization; Transportation and Grid Applications; Estimation and Identification Methods, Tracking, Detection, Alternative Propulsion Systems; Ground and Space Vehicle Dynamics; Intelligent Transportation Systems and Control; Energy Harvesting; Modeling and Control for Thermo-Fluid Applications, IC Engines, Manufacturing
  • San Antonio, Texas, USA, October 22–24, 2014
  • Conference Sponsors: Dynamic Systems and Control Division
  • ISBN: 978-0-7918-4619-3
  • Copyright © 2014 by ASME

abstract

For recognizing human motion intent, electromyogram (EMG) based pattern recognition approaches have been studied for many years. A number of methods for classifying EMG patterns have been introduced in the literature. On the purpose of selecting the best performing method for the practical application, this paper compares EMG pattern recognition methods in terms of motion type, feature extraction, dimension reduction, and classification algorithm. Also, for more usability of this research, hand and finger EMG motion data set which had been published online was used. Time-domain, empirical mode decomposition, discrete wavelet transform, and wavelet packet transform were adopted as the feature extraction. Three cases, such as no dimension reduction, principal component analysis (PCA), and linear discriminant analysis (LDA), were compared. Six classification algorithms were also compared: naïve Bayes, k-nearest neighbor, quadratic discriminant analysis, support vector machine, multi-layer perceptron, and extreme machine learning. The performance of each case was estimated by three perspectives: classification accuracy, train time, and test (prediction) time. From the experimental results, the time-domain feature set and LDA were required for the highest classification accuracy. Fast train time and test time are dependent on the classification methods.

Copyright © 2014 by ASME

Figures

Tables

Interactive Graphics

Video

Country-Specific Mortality and Growth Failure in Infancy and Yound Children and Association With Material Stature

Use interactive graphics and maps to view and sort country-specific infant and early dhildhood mortality and growth failure data and their association with maternal

NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In