Fundus Image Tracking via Pattern Matching for Eye Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current ophthalmologic imaging techniques face challenges in accurately selecting a fundus image that is less affected by eye motion, leading to increased burdens on operators and patients due to the need for repeated template image extraction and inefficient movement detection.
Innovation Solution
An ophthalmologic apparatus and control method that extracts a template image from a first fundus image and evaluates subsequent images using pattern matching to determine if they are affected by eye motion, allowing for the selection of a fundus image with minimal eye motion and reducing the need for repeated image acquisition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If pattern matching is executed without determining whether the fundus image is proper, then the process is simple, but template image extraction requires repetition increasing time and operator burden
Solution Approach 1:
The patent applies preliminary action by determining whether a fundus image is proper (free from eye motion) before extracting the template image. This preliminary check prevents the need for repeated extraction later, resolving the contradiction between process simplicity and time efficiency.
2Measurement precision
If additional hardware is added for detecting eye movement, then tracking accuracy improves, but device complexity increases
Solution Approach 1:
The patent applies self-service by using the fundus image itself to detect eye motion without requiring additional dedicated hardware. The system evaluates the fundus image to determine if it is proper, making the image serve dual purposes for both diagnosis and motion detection, thus resolving the contradiction between accuracy and complexity.
3Productivity
If template image is extracted from improper fundus image, then extraction is quick, but matching accuracy decreases leading to false detection
Solution Approach 1:
The patent applies preliminary action by evaluating the fundus image to determine if it is proper before extracting the template. This preliminary evaluation ensures that only images without eye motion are used for template extraction, guaranteeing matching accuracy while maintaining efficiency through automated selection.
Data Source
Figure 1
Figure 2
Figure 3A~3C
AI summary
There is provided an ophthalmologic apparatus having a tracking function that can select a fundus image that is less affected by eye motion to reduce burdens on an operator/a patient in fundus imaging, wherein the ophthalmologic apparatus picks up a first fundus image (202), extracts a characteristic point as a template from the first fundus image (203), executes pattern matching on a second fundus image (205) to determine presence/absence of eye motion, and decides a tracking template image (207).