Stereo Camera Gaze Detection for Fundus Imaging Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional fundus imaging techniques face challenges in accurately and efficiently aligning the imaging path of a fundus camera with a human subject's pupil and in consistently determining the subject's gaze angle, leading to potential misalignments and reduced image quality.
Innovation Solution
A fundus camera system that includes a stereo camera for detecting the pupil and determining the gaze angle, using machine-learning models to analyze stereo images and automatically align the imaging path with the pupil, and to verify the gaze angle before capturing a fundus image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional fundus imaging techniques are used, then the imaging process is simple, but the alignment accuracy between imaging path and pupil is poor
Solution Approach 1:
The system performs preliminary pupil detection and gaze angle determination using stereo cameras and machine learning models before initiating the fundus imaging process. This preliminary alignment preparation ensures that the imaging path is pre-positioned accurately at the pupil location, resolving the alignment accuracy issue while maintaining a structured workflow
Solution Approach 2:
The patent introduces stereo cameras and machine learning-based gaze tracking systems as intermediary components between the operator and the fundus camera. These intermediaries automatically detect pupil position and gaze angle, providing precise alignment data without requiring manual intervention, thus improving measurement precision while managing system complexity through modular architecture
2Manufacturing precision
If manual alignment methods are used, then the operation process is simple, but the image quality is reduced
Solution Approach 1:
The system performs self-alignment by automatically detecting the pupil position and gaze angle using stereo cameras and machine learning models. The fundus camera then autonomously positions the imaging path at the pupil without requiring manual adjustment by the operator, ensuring consistent image quality while reducing operator intervention to essential steps only
Solution Approach 2:
The system implements feedback loops where stereo cameras continuously monitor pupil position and gaze angle, and machine learning models process this data to provide real-time alignment corrections. This feedback mechanism ensures that the imaging path remains accurately positioned at the pupil throughout the imaging process, maintaining high image quality while minimizing the need for manual operator adjustments
3Productivity
If automated pupil detection is implemented, then the alignment efficiency is improved, but the device complexity increases
Solution Approach 1:
The patent replaces manual mechanical alignment operations with automated optical detection systems. Stereo cameras capture pupil images, and machine learning models automatically determine gaze angles and alignment parameters, substituting the mechanical adjustment process with computational methods that improve alignment efficiency while managing complexity through software-based solutions
Solution Approach 2:
The stereo camera system serves multiple functions: it detects pupil position, determines gaze angle, and provides alignment verification. This multi-functionality improves alignment efficiency by consolidating multiple detection tasks into a single integrated system, thereby managing device complexity through versatile components rather than separate specialized devices
Data Source
AI summary
The techniques described herein relate to systems, apparatus, articles of manufacture, and methods for gaze angle triggered fundus imaging. An example method includes detecting a pupil of a subject in a stereo image, controlling at least one actuator to align an imaging path of a fundus camera with the pupil after detecting the pupil in the stereo image, and capturing a fundus image of a retina of the subject at a gaze angle after determining that the pupil is oriented towards a target direction based on the gaze angle associated with the pupil in the image.