A novel electrooculogram-based human computer interface and its application as a virtual keyboard

Abstract

In this study, the design strategy of a novel electrooculugram (EOG) based human computer interface (HCI) and its application are presented. Due to eye movements electric potentials are generated accross the cornea and retina. This electrical potentials are the source of the EOG signal. To establish a new channel of communication between patients having motor neuron problems and their envioriment is important to make their life easy. Using EOG for HCI is more efficient than electroencephalogram based methods for patients who are still able to move their eyes. The novel system is microcontroller based and has 85 dB over-all common mode rejection ratio, 0.6 µV(p-p) input-referred noise and 176 Hz sampling rate. The subject can write a 5 letter word in 25 seconds. Nearest neighborhood method is used for classification and its performance is 92%.