Related Experiment Video
Updated: Jun 23, 2025
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
434
Electromyography signal based hand gesture classification system using Hilbert Huang transform and deep neural
Mary Vasanthi S1, Haiter Lenin A2, Yasser Fouad3
1Department of Electronics and Communication Engineering, St Xavier's Catholic College of Engineering, Nagercoil, Tamilnadu, India.
Heliyon
|June 24, 2024
Summary
This study introduces an advanced hand gesture recognition system using surface electromyography (sEMG) and deep neural networks (DNNs). The novel approach achieves a high 98.5% accuracy for prosthetic hand control.
Area of Science:
- Biomedical Engineering
- Neuroscience
- Artificial Intelligence
Background:
- Current methods for hand gesture recognition, such as data gloves and vision-based systems, have limitations including user inconvenience and high costs.
- Surface electromyography (sEMG) signals, reflecting muscle activity, offer a promising alternative for controlling prosthetic devices.
- Developing accurate and efficient methods for classifying sEMG data is crucial for advancing bio-control systems.
Purpose of the Study:
- To develop an automated hand gesture recognition system using electromyography (sEMG) to overcome limitations of existing technologies.
- To enhance the performance of hand gesture recognition systems through the application of artificial classifiers, specifically Deep Neural Networks (DNNs).
- To improve classification accuracy for prosthetic hand control.
Main Methods:
- Utilized surface electromyography (sEMG) data to capture hand muscle activation patterns.
- Employed signal processing techniques, including the Hilbert Huang Transform (HHT), to extract essential signal features.
- Developed and trained a Deep Neural Network (DNN) classifier using the extracted sEMG features for hand gesture recognition.
Main Results:
- The proposed DNN-based system achieved a high classification accuracy rate of 98.5%.
- The integrated approach of HHT feature extraction and DNN classification significantly outperformed alternative methods.
- Demonstrated the effectiveness of sEMG-based gesture recognition for potential application in prosthetic hands.
Conclusions:
- The developed sEMG-based hand gesture recognition system provides a robust and accurate method for controlling prosthetic devices.
- The combination of advanced signal processing (HHT) and Deep Neural Networks offers a powerful framework for bio-control applications.
- This research lays the groundwork for more intuitive and effective human-computer interaction and upper limb prostheses.
