Research Status: Completed
Researcher: Sunu Wibirama, S.T., M.Eng., IPM.
This research develops a three-dimensional eye-tracking system based on dual cameras to detect vertigo more accurately and practically. The system is mounted on lightweight glasses equipped with two mini infrared cameras to capture stereo images of a single eye. By utilizing center-of-mass algorithms, template matching, and direct linear transformation (DLT), the system can precisely track the pupil center and iris patterns, as well as reconstruct eye position in 3D without requiring complex per-patient calibration or camera adjustments.
The testing results show very low angular errors: 0.15° in the horizontal direction, 0.14° in the vertical direction, and 0.20° in the torsional direction—outperforming other methods. The system also provides real-time 3D visualization of the eyeball using OpenGL technology.
In further testing, the system is capable of tracking both voluntary eye movements, such as focusing on moving objects, and involuntary movements, such as nystagmus in vertigo patients.
This system is affordable, easy to use, and accurate, opening opportunities for broader applications of 3D eye-tracking in clinical settings. With further development, it can be adapted for high-speed or wireless applications, expanding its potential in diagnosing and understanding eye and vestibular system disorders
