EMG pattern recognition by neural networks for multi fingers control

Key Points

• Recognition of finger movement and joint angles was achieved using EMG signals and neural networks, enhancing robotic control.
• Key metrics include classification of movements, integrating data from 5 fingers and 10 joints into a robotic hand design.
• Assessment involved online simulations and experiments focused on real-time interpretations of EMG patterns.
• This implies potential improvements in the design of intuitive robotic systems; further applications may follow.