Eye-gaze detection based on the pupil detection technique using two light sources and the image difference method

Abstract

A new eye-gaze detection method was developed to use on a support system for the physically disadvantaged, such as amyotropic lateral sclerosis patients. This method was based on the pupil detection technique using two light sources and the image difference method. In this paper, hardware which detects the pupil and glint images (feature points for determining the eye-gaze) was developed. When wearing eye glasses, the system consisted of the image differentiator, the pupil detector including the noise reducer (dilation and erosion operations) and edge detector, the two circular moving window processors, the centroid detector and the glint edge detector. By using this system, the eye-gaze was detected even when the eye glass reflected light was seen in the images.