Real Time Classification for Robotic Arm Control Based Electromyographic Signal

Bioengineering developments have enabled in increasingly advanced prosthetic devices for amputees and paraplegic people. With the advancement of technology for control and the enhancement of people's living conditions, the demand for high-quality prostheses among disabled persons has grown significantly. Real-time characterization of bio signals, such as electromyographic (EMG) data acquired from intact muscles, is required for control of such devices. This study provides a control of robotic arm depend on the surface EMG of the human arm. Our system's unique characteristics include physiologically informed forearm muscle selection for EMG signal recording, simple feature extraction from EMG signals and clever hand gesture selection for easy classification. The Arduino Uno is used to operate the servo motors that move our model, and it handles the EMG signals that are produced at varying levels of recurrence and concentration. EMG signals from the forearm surface of eight healthy individuals were collected during the experiments. The outcomes of the experiment show that the system gesture real - time tracking rate is achieved up to 90%, and it responds fast, laying the groundwork for future artificial limb control.