High-precision eye-tracking system using Purkinje reflections for real-time gaze localization in virtual and augmented reality applications
Institute Reference: 1-18137
Eye-tracking systems are widely used in fields like clinical diagnostics, human-computer interaction, and virtual/augmented reality. Traditional commercial eye-trackers often lack the precision required for advanced applications, providing only coarse gaze localization. Dual-Purkinje Image (DPI) tracking offers high precision but has been limited to specialized labs due to its complexity and cost. The market demands a simpler, more accessible, and high-resolution eye-tracking system for broader applications.
Researchers at the University have developed a high-resolution, real-time eye-tracking system leveraging a combination of Purkinje reflections (images formed by light reflecting off different layers of the eye) and advanced computational algorithms. The system includes an eye-tracking device with an infrared illuminator, a camera or imager, and a processor. It tracks the movement of the eye by capturing and analyzing at least two Purkinje reflections to determine gaze location. The device is designed for both head-mounted displays (HMDs) like VR and AR systems and can achieve high precision with minimal moving parts.
The method also employs predictive algorithms to handle potential tracking inaccuracies caused by eye movements and blinking, allowing for reliable and low-latency performance even at low frame rates. Calibration processes ensure individual adaptation, enhancing tracking accuracy across various applications.
Available for industry sponsored research and licensing.