Investigation on the Sampling Frequency and Channel Number for Force Myography Based Hand Gesture Recognition

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction

Background:

• Force myography (FMG) indirectly measures muscle contractions using pressure sensors, serving as a viable alternative to traditional myoelectric signals for hand gesture recognition.
• Cost-effective gesture recognition necessitates understanding the minimum required sampling frequency and number of channels for FMG systems.

Purpose of the Study:

• To investigate the impact of sampling frequency and channel count on the accuracy of FMG-based hand gesture recognition.
• To identify the minimum FMG signal parameters for accurate gesture recognition.

Main Methods:

• A 16-channel hardware system was developed to capture forearm FMG signals at a maximum sampling frequency of 1 kHz.
• A database of FMG signals from 10 subjects was created to evaluate gesture recognition accuracy under varying conditions.

Main Results:

• High accuracy (up to 99%) was achieved with a 1 kHz sampling rate and 16 channels using four out of five classifiers.
• Gesture recognition accuracy remained high with a sampling frequency as low as 5 Hz for static movements.
• Reducing the channel count significantly impacted accuracy, with eight channels suggested for optimal gesture recognition.

Conclusions:

• FMG is a promising, cost-effective technology for hand gesture recognition.
• Optimizing sensor placement and reducing channel count to eight can maintain high accuracy while lowering system costs.
• The study provides crucial insights into parameter optimization for practical FMG applications.