Abstract:
This paper discusses a method of implementing the detection of pupil centre and corneal reflection for a low
cost, remote eye tracking system that is designed for usability analysis researches. The method we have used to
measure the eye gaze is tracking corneal reflection and the pupil centre of the eye, where the paper focuses on
how those features are tracked from a video input that have captured the subjects ' eye movements. The method
suggested here is capable of tracking the pupil centre even from an eye with less colour contrast between the
pupil and the iris and also endures head movements of the subjects. The test results are included in terms of
accuracy of the method to the changes in point of gaze and distance between the computer screen and the user.
The limitations of modern eye trackers and possible solutions are also addressed in this paper.
