Robust Pupil Detection in the Presence of Eyeglasses

Abstract

Eye movements are known to be deeply related to a person's emotional state. Gaze tracking is a contactless method to detect these, commonly using the spatial relationship between the pupil and corneal reflection. However, it does not perform robustly when the user is wearing eyeglasses since light reflected from the surroundings changes the appearance of the pupil. In this research we propose and evaluate a pupil detection method that can perform robustly even in the presence of such reflection.