Sensors play a vital role in health care applications. An artificial hand, based on sensors is gaining importance. Artificial hand is designed in such a way that its fabrication is kept simple and durable for smooth functioning of desired actions.
The artificial hand helps the user to perform activities such as pick, grab, pinch, and gestures such as point, ok, and cool sign. It uses Electro-Myo-Graphic (EMG) signals generated by the amputee’s muscle movements which are captured through the EMG electrodes. Two muscles; bicep and tricep are identified from the upper part of the amputee. The EMG electrodes are connected to capture signals from the muscles. The captured signals are fed to the EMG module, processed, filtered and gain is added.
A machine learning technique is implemented in Arduino. The captured signals received by the Arduino are provided to the machine for learning algorithm. The test signals are then provided to Arduino and filtered to perform different activities like grab, pick and pinch and gestures (ok, point and cool signs).
The experiment results show that the artificial hand works efficiently with an accuracy ranges from 90 to 95 percent. It is observed that the accuracy of filtering the signals into grab and pick remain high as compared to pinch, point, cool and ok signs.