Retinal Image Gaze Tracking Less Sensitive to Camera Movement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gaze tracking technologies are limited by image quality, pupil size variations, and sensitivity to camera position, leading to errors in gaze estimation, especially for distant targets, and are not suitable for real-time applications.
Innovation Solution
A system and method for gaze tracking using a camera positioned to image the retina, calculating a change in eye orientation based on comparisons between images, and determining gaze direction by finding spatial transformations or fovea location, less sensitive to camera movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If video-based eye tracking uses corneal reflection and pupil center as features, then gaze tracking can be performed, but measurement precision deteriorates due to sensitivity to camera position and pupil size variations
Solution Approach 1:
The patent creates a digital reconstructed image of the retina by combining multiple retinal retroreflections (RRs). Instead of directly observing the eye features that are sensitive to camera position, the system copies the retinal structure through RR imaging and analyzes this reconstructed representation to determine gaze direction, thereby eliminating the sensitivity to camera slippage and pupil size variations
Solution Approach 2:
The patent introduces retinal retroreflections as an intermediary mechanism between the camera and the eye. By capturing RRs and combining them to form a digital retinal image, the system creates an intermediate representation that is independent of camera position, serving as a mediator that enables accurate gaze measurement without direct dependence on camera-eye geometric relationships
2Measurement precision
If retinal imaging is used for gaze tracking, then accuracy is improved, but device complexity increases due to the need for multiple images and spatial transformation calculations
Solution Approach 1:
The patent performs preliminary action by capturing multiple retinal retroreflections and combining them to create a digital reconstructed retinal image before gaze analysis is required. This pre-processing step establishes a reference retinal structure that simplifies subsequent gaze direction calculations by providing a stable baseline for comparison
Solution Approach 2:
The patent segments the retinal imaging process into distinct components: capturing individual retinal retroreflections, combining these RRs to form a digital retinal image, and then analyzing this segmented representation for gaze direction. This segmentation allows each step to be optimized independently and reduces overall computational complexity
3Device complexity
If single RR analysis is used, then device complexity is reduced, but measurement precision deteriorates because any individual RR may result from more than one gaze direction
Solution Approach 1:
The patent merges multiple retinal retroreflections into a single digital reconstructed retinal image. By combining the information from multiple RRs, the system creates a comprehensive retinal representation that maintains unique correspondence with gaze direction, eliminating the ambiguity present in single RR analysis while preserving computational efficiency
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
Provides accurate and real-time gaze tracking with reduced errors due to camera position changes, enabling applications in human-machine interaction, virtual reality, and biometric identification.
Implementation Method 1
a camera lens configured to focus light originating from the person's retina on the image sensor
Implementation Method 2
a light source producing light emanating from a location near the retinal camera
Data Source
AI summary
A system and method for inputting to a statistical model a reference image, which is associated with a known direction of gaze of a person and which includes at least a portion of the person's retina, and an input image which is associated with an unknown direction of gaze of the person and which includes at least a portion of the person's retina. The statistical model is trained on multiple images of portions of retinas obtained at known directions of gaze and can output an estimation of a change in orientation of an eye of the person. A signal generated based on the estimation can be used to control a device.