Gesture recognition using real time EMG

Abstract

Electromyography shows the electrical activity of the muscles of the body. EMG signals can be recorded using surface electrodes. The SEMG signals thus obtained are used for varied applications most of which contribute towards prosthetic technology and rehabilitation engineering. SEMG signals are used as control signals for robotic arm functions as well. The aim of this study is to construct a prototype which uses these SEMG signals recorded form the brachioradialis muscle of the forearm to control the movement of powerpoint slides transmitted wirelessly in real time. The methodology involves using surface electrodes to acquire EMG signals which is sent to an ADC interfaced with a microcontroller. Using Zigbee protocol the signal is transmitted wirelessly to a PC where the powerpoint slides are contained. The powerpoint is displayed in visual basic window and embedded C language is used for microcontroller programming. The results showed 100mV voltage threshold for opening and closing of the fingers in the EMG signal. For the operation of the slides the fast upward action of the forearm caused a voltage higher than 90mV which leads to the backward movement of the slide and slow upward action of the forearm caused a voltage range between 85mV–89mV which leads to the forward movement of the slide. When this voltage is matched in the database the slides move. In conclusion a wireless control using EMG in real time application is achieved for operation.