Fast and Accurate Pupil Positioning Algorithm using Circular Hough Transform and Gray Projection
Milad Soltany, Saeid Toosi Zadeh, Hamid Reza Pourreza · 2011
A fast pupil-positioning algorithm for real-time eye tracking is proposed in this paper. It is significant accurately locate the pupil positi2on in an eye tracking system. Commonly used method is combining edge detection algorithms and ellipse fitting. Edge detection algorithms are used to detect edges of the pupil while the ellipse fitting is supposed to find the optimum ellipse that finely fits the pupil and the centre of the ellipse is regarded as the location of the pupil. This algorithm is acceptable except that the definition of the pupil edge has a great influence on its efficiency and ellipse fitting is a time consuming method. This paper focuses on accuracy of the primary algorithm and some improvement on it and uses circular Hough transform to detect pupil area. Firstly, mainly localize the pupil position by gray projection. Secondly, fit a circle to pupil using circular Hough transform.