Abstract:

Rehabilitation for individuals with disabilities often requires precise and repetitive hand movements to restore functionality. Traditional methods can fall short in providing real-time feedback and adapting to the unique needs of each individual. Surface electromyography (sEMG) signals, which measure muscle electrical activity, present a promising solution for improving the recognition and interpretation of hand gestures.

This research project aims to revolutionize rehabilitation learning for individuals with disabilities by achieving real-time recognition of hand gestures using sEMG signals. By applying a meta-learning approach, this study seeks to accurately identify distinct hand gesture patterns, thereby enhancing the effectiveness and personalization of rehabilitation programs.

The research investigates the detection accuracy of various hand gesture patterns, including resting hand, fisting hand, wave-like finger motions, and wrist rotations. These gestures are processed using continuous wavelet transform and classified by the Reptile algorithm. The identified gestures are then utilized to control a robotic vehicle implemented on the Arduino platform, with Bluetooth communication ensuring seamless operation.

Significant improvements in system performance were achieved through the integration of multiple sEMG channels and optimization of time windows, resulting in enhanced accuracy and reduced latency. Preliminary evaluations of the system demonstrated its ability to recognize single instances of hand gestures with a latency of approximately three seconds. The findings suggest that this intuitive and responsive control system holds substantial potential for personalized rehabilitation, providing an engaging and effective tool for motor skill recovery.

This project highlights the promising intersection of meta-learning and assistive technology, paving the way for future advancements in the field of rehabilitation. The innovative approach and encouraging results indicate a significant step forward in creating more adaptive and responsive rehabilitation tools for individuals with disabilities.

Keywords:

Rehabilitation learning, personalized rehabilitation, hand gesture recognition, assistive technology, sEMG signals.