Eye Profiling Through 3D Corneal Mapping for Accurate Gaze Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye tracking systems suffer from inaccuracies in determining gaze direction due to assumptions about the shape and position of eye features, particularly the cornea, leading to errors in estimating pupil and corneal reflections.
Innovation Solution
An eye tracking system that utilizes multiple light sources to capture corneal reflections in different illuminated sub-regions, adjusts a corneal surface model to minimize errors in ray tracing paths, and determines three-dimensional positions of these reflections to generate a personalized topographical profile of the cornea, improving gaze estimation accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If standard corneal models are used for gaze estimation, then the system complexity is low, but the measurement precision of gaze direction deteriorates due to assumptions about corneal shape and position
Solution Approach 1:
The system performs preliminary corneal topography mapping by capturing multiple images of the corneal surface with different illuminated sub-regions before gaze estimation. This preliminary action creates an accurate 3D representation of the individual's corneal surface, which is then used to improve subsequent gaze measurements without requiring complex real-time processing during actual gaze tracking
Solution Approach 2:
The system transitions from 2D image analysis to 3D corneal surface modeling by determining three-dimensional positions of corneal reflections and constructing a topographical profile. This dimensional enhancement allows for more accurate gaze estimation by accounting for the actual 3D geometry of the individual's cornea rather than relying on simplified 2D assumptions
2Measurement precision
If multiple light sources are used to capture different illuminated sub-regions, then the topographical profile accuracy improves, but the device complexity increases
Solution Approach 1:
The corneal surface is divided into multiple illuminated sub-regions, each captured by a separate light source. By segmenting the corneal surface into different zones and illuminating them sequentially or simultaneously with multiple light sources, the system achieves comprehensive 3D mapping of the entire corneal surface with high precision
Solution Approach 2:
The system uses periodic illumination patterns where multiple light sources are activated in a structured sequence to illuminate different sub-regions of the cornea. This periodic action allows systematic coverage of the entire corneal surface while maintaining manageable system complexity through controlled timing and coordination of light source activation
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
The system provides enhanced accuracy in determining the corneal center and gaze direction by accounting for individual corneal variations, thereby improving the precision of gaze estimation and reducing errors associated with standard corneal models.
Implementation Method 1
each image comprising a different illuminated sub-region of a corneal surface of the eye, the illuminated sub-region comprising: a first corneal reflection of a first light source from the illuminated sub-region; and a second corneal reflection of a second light source from the illuminated sub-region
Data Source
AI summary
An eye tracking system comprising a processor configured to: receive a plurality of images of an eye from a camera, each image comprising an illuminated sub-region of a cornea of the eye, the illuminated sub-region comprising: a first corneal reflection of a first light source from a surface of the illuminated sub-region; and a second corneal reflection of a second light source from the surface of the illuminated sub-region; for each image, determine a three-dimensional, 3D, position of the first corneal reflection and a 3D position of the second corneal reflection based on a position of the camera, a position of the first light source, a position of the second light source, positions of the first and second corneal reflections in the image and a corneal surface model for the illuminated sub region; and output a topographical profile of the cornea based on the 3D positions determined for each image.