Calibration-Free Gaze Tracking Using Glint-Pupil Displacement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye gaze tracking methods are inaccurate due to reliance on head orientation and require cumbersome calibration, and are intrusive or constrained by the need for static head positions.
Innovation Solution
A video-based gaze tracking method using a single camera to focus on the eye, sampling pupil and glint images to estimate gaze points with parameters Δx, Δy, r, θ, gx, and gy, allowing for natural head movement and calibration-free operation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If video-based gaze tracking methods are used, then the subject comfort is improved, but the accuracy is worsened
Solution Approach 1:
The method segments the gaze tracking task into multiple measurement components: pupil center position, glint center position, and their relative displacement. By separately measuring these elements and combining them, the system achieves both comfort (video-based) and accuracy (multiple measurement points).
Solution Approach 2:
The glint serves as an intermediary reference point that mediates between the camera view and the pupil position. The glint's known geometric relationship to the light source and cornea provides a stable reference frame that enables accurate gaze calculation from video images alone.
2Device complexity
If existing gaze tracking methods that relate gaze to head orientation are used, then the setup is simplified, but the accuracy is worsened
Solution Approach 1:
The method extracts the glint position as a separate measurable quantity from the video image, independent of head orientation assumptions. This extracted glint position serves as a direct geometric reference that eliminates the need for complex head orientation modeling while maintaining simplicity.
3Measurement precision
If calibration process is performed for each individual, then the accuracy is improved, but the time consumption is worsened
Solution Approach 1:
The system performs self-calibration by automatically determining the glint center position and pupil center position from video images without requiring manual marking or adjustment. The geometric relationships are automatically computed from the captured images, eliminating time-consuming manual calibration procedures.
4Measurement precision
If static head constraint is imposed, then the gaze tracking accuracy is improved, but the ease of operation is worsened
Solution Approach 1:
The method dynamically adapts to head movement by continuously tracking both the pupil and glint positions in the video image sequence. The glint position automatically adjusts with head orientation changes, providing a moving reference frame that maintains accuracy regardless of head position or motion.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Accurately tracks gaze points without calibration, enabling robust and unobtrusive eye movement tracking even with natural head movements, improving accuracy and user comfort.
Implementation Method 1
sampling eye gaze data pertaining to a glint and pupil image of the eye in an image plane of the single camera
Implementation Method 2
directing light toward the eye
Data Source
AI summary
A method and computer system for tracking eye gaze. A camera is focused on an eye of subject viewing a gaze point on a screen while directing light toward the eye. Eye gaze data pertaining to a glint and pupil image of the eye in an image plane of the camera is sampled. Eye gaze parameters are determined from the eye gaze data. The determined eye gaze parameters include: orthogonal projections of a pupil-glint displacement vector, a ratio of a major semi-axis dimension to a minor semi-axis dimension of an ellipse that is fitted to the pupil image in the image plane, an angular orientation of the major semi-axis dimension in the image plane, and mutually orthogonal coordinates of the center of the glint in the image plane. The gaze point is estimated from the eye gaze parameters.