Classification of Angle of Eye Movement using Electrooculography with LED testing sphere
Tasawan Puttasakul, Jiranun Auttakunchai, Phichet Kawilo · 2022 International Electrical Engineering Congress (iEECON) · 2022
This paper proposed the classification of the direction and degree of eye movements using EOG signals 2 channels based on discrete cosine transform and power of discrete signal. The EOG signal collect from the electrode placed on the around of eye in vertical and horizontal and using LED testing sphere to guide the EOG signal from the eye muscles. The k-means Clustering used to identify the posture and degree of eye view. The results show that this method can identify four motion gestures: up sight, down sight, left sight and right sight, with an average accuracy of 98.57% and can distinguish the five degrees of view: 5°, 10°, 15°, 20° and 25° with an average accuracy of 91.94%.