Pupil Detection Methods for Eye Tracking

Abstract

The theory that pupillary changes can reflect human psychophysical character was confirmed in several conducted studies. Authors of those studies have used devices called eye trackers in order to measure pupillary changes, blinking rate, eye movement patterns or even a gaze position. The main goal of those apparatuses is to locate the pupil on a given image using implemented pupil detection algorithms. In this paper the main focus will be set on several states of art pupil detection methods. All represent different approaches to performing the same tasks in difficult situations for ex.: low/high condition, low image resolutions or low contrast.