How to use one surface electromyography sensor to recognize six hand movements for a mechanical hand in real time: a...

Area of Science:

• Biomedical Engineering
• Human-Computer Interaction
• Signal Processing

Background:

• Surface electromyography (sEMG) signals reflect muscle activity and force.
• Existing human motion intention recognition methods often require multiple sensors, specific placement, and high-end hardware.
• There is a need for more accessible and efficient methods for recognizing multiple movements.

Purpose of the Study:

• To develop a real-time motion intention recognition system using a single sEMG sensor.
• To encode sEMG signals using Morse code for movement recognition.
• To apply this method for controlling a mechanical hand with high accuracy and low hardware demands.

Main Methods:

• Collected short-time and long-term muscle contraction signals using a single sEMG sensor.
• Encoded extracted sEMG signals using the Morse code method.
• Developed a mapping system to translate Morse code into six distinct hand movements for recognition.

Main Results:

• Achieved an average hand movement recognition accuracy of 94.87% ± 2.36%.
• Demonstrated an average subject adjustment time of 34.89 seconds.
• Reported a single movement execution time of 381 milliseconds.
• Showcased robustness to variations in sensor placement.

Conclusions:

• The proposed Morse code-based sEMG recognition method enables single-sensor control of multiple movements.
• This approach offers high accuracy, low hardware requirements, and is insensitive to sensor placement differences.
• The method presents a significant advancement for practical human-computer interfaces and prosthetic control.