Related Experiment Video
Updated: Jul 10, 2026
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
Introducing a new multi-wavelet function suitable for sEMG signal to identify hand motion commands
1Department of Electrical Engineering, Sharif University of Technology, Biomedical Engineering and Robotic Laboratories, Tehran, Iran. mahdi_khezri_ee@yahoo.com
Summary
A new multi-wavelet function improves hand movement recognition from surface electromyogram (sEMG) signals. This novel approach achieved 87% accuracy, surpassing existing mother wavelets by 9%.
Area of Science:
- Biomedical Engineering
- Signal Processing
- Machine Learning
Background:
- Electromyogram (EMG) signal feature selection using wavelet transforms is crucial for human-computer interaction.
- Surface EMG (sEMG) analysis is vital for understanding and recognizing hand movements.
- Accurate feature extraction from sEMG is challenging due to signal complexity.
Purpose of the Study:
- To introduce a novel multi-wavelet function for enhanced sEMG feature selection.
- To improve the accuracy of hand movement recognition using sEMG signals.
- To evaluate the performance of the proposed multi-wavelet against traditional mother wavelets.
Main Methods:
- Development of a new multi-wavelet function by integrating established mother wavelets.
- Application of the proposed multi-wavelet to sEMG signals from eight distinct hand motion classes.
- Utilizing Discrete Wavelet Transform (DWT) features, including local extrema and zero crossings (ZC).
Main Results:
- The proposed multi-wavelet function demonstrated superior performance in sEMG feature extraction.
- Hand movement recognition accuracy reached 87% using the novel multi-wavelet.
- This represents a significant improvement over the best-performing traditional mother wavelet, which achieved 78% accuracy.
Conclusions:
- The developed multi-wavelet function effectively captures sEMG signal characteristics for improved recognition.
- This new approach offers a promising advancement for accurate hand movement classification systems.
- The findings highlight the potential of integrated wavelet functions in biomedical signal processing.
