Using structured illumination to enhance video-based eye tracking

Video-based eye tracking commonly relies on the positional difference between the pupil center and the first-surface corneal reflection of a light source (CR) to map the line of sight. In this paper, a new structured illumination configuration, utilizing an array of IREDs to illuminate the eye, is proposed to address two common problems in current systems: 1) loss of CR during extreme eye movements, and 2) specular reflection artifacts being mistaken for the desired CR. The design provides reliable difference vectors even during extreme eye movements. The challenge of spurious specular reflections is solved by a two-stage processing approach. First, potential CRs and the pupil are isolated based on statistical information in an eye image. Second, genuine CRs are distinguished by a novel CR location prediction technique. A prototype implementation in the RIT Lightweight Eye Tracker [1] is described.