Implementation of sEMG-based robotic hand

This paper describes utilization of surface electromyography (sEMG) identification for driving robotic hand. Therefore, this implementation will give other alternatives to the stroke survivors or paralysis patients in order to help their activities. This system detecs the movement of two fingers (thumb and index finger) by using threshold as pattern recognition algorithm. This pattern recognition generates a decision that will translate into command to drive the robotic hand. The system using two differential inputs ModularEEG. It has 256Hz as frequency sampling. This research uses BrainBay as an application in order to pre-process sEMG signal, process and implement the pattern recognition algorithm. As a result, system gives time response about 232 ms and 86,2% accuracy in order to distiguish tumb and index finger.