[Research on Exact Location Algorithm of Pupil Center Based on Ellipse Fitting]

Area of Science:

• Biomedical Engineering
• Computer Vision
• Ophthalmology

Background:

• Precise eye tracking is crucial for clinical applications.
• Existing pupil detection methods can be susceptible to artifacts like glare and squint.
• Real-time and accurate pupil localization is a persistent challenge.

Purpose of the Study:

• To develop and validate a robust pupil center location algorithm for clinical eye tracking.
• To improve the accuracy and reliability of pupil detection under challenging conditions.
• To enable fast and real-time pupil center identification.

Main Methods:

• An eye image was captured using infrared light and preprocessed to obtain a binary image.
• Pupil candidates were roughly filtered based on contour pixel count.
• Ellipse fitting using the least squares method was performed on filtered contours.
• The pupil center was identified as the ellipse center minimizing the variance of distance to the contour.

Main Results:

• The proposed algorithm demonstrated accurate pupil center identification.
• Robust performance was observed even in the presence of white spot interference and squint.
• The algorithm achieved quick and real-time pupil localization.

Conclusions:

• The least squares-based pupil center location algorithm provides a precise and reliable solution for clinical eye tracking.
• This method offers significant advantages in handling image artifacts common in clinical eye tracking.
• The algorithm's efficiency supports real-time applications in various medical fields.