A Pupil Tracking Method Based on Particle Filter

Eye tracking technology has attracted more and more attention at home and abroad, becomes a hot topic in many disciplines. The part pupil tracking is very important. As the primary means of dealing with nonlinear and non-Gaussian filtering problem, Particle Filter has overcame the Kalman Filters's defects, it has no distribution limit to process noise and observation noise. While SIR (Sequential Importance Resampling) sampling method can eliminate particle degradation in the traditional particle filter. In this paper, we track the pupil by Particle Filter based on SIR. Experiments show that the method can be accurate and real-time to track the pupil, and it has high value of research and application.