Corneal Vertex Detection Using Calibrated Triangulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining the corneal vertex position in space are imprecise and limited by the need for significant overlap between camera images, restricting camera arrangement and making it difficult to accurately calculate centering parameters like pupillary distance and corneal vertex distance.
Innovation Solution
A computer-implemented method using calibrated images from cameras with known extrinsic and intrinsic properties, employing triangulation and correction calculations to determine the corneal vertex position, allowing for non-overlapping camera angles and improving precision through pupil-based or reflex-based evaluations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If stereo algorithms based on finding correspondences in images are used to determine corneal vertex position, then the position can be detected, but a large overlap of images is required which severely limits camera arrangement choices
Solution Approach 1:
The patent introduces a light source as an intermediary element that projects a light pattern (such as a crosshair or grid) onto the cornea. This light pattern serves as a mediator that can be detected by cameras from different angles, enabling corneal vertex detection without requiring significant image overlap. The light pattern provides reference points that facilitate correspondence finding between images taken from different camera positions.
Solution Approach 2:
The patent changes the approach from direct corneal feature detection to detecting the position of a light pattern projected onto the cornea. By projecting a known geometric pattern (crosshair, grid, or other shapes) and tracking its position across multiple images, the system can determine corneal vertex location without requiring the cameras to have overlapping fields of view, thus enabling more flexible camera arrangements.
2Measurement precision
If manual annotation of corneal vertex is performed, then opticians can determine position, but the process is time-consuming and prone to human error
Solution Approach 1:
The system enables automatic detection of the corneal vertex position through computer vision algorithms that analyze the light pattern projected onto the cornea. The detection process is self-service in the sense that the system automatically identifies the corneal vertex location without requiring manual annotation by opticians. The algorithm processes the images and calculates the position autonomously, eliminating human intervention in the measurement process.
Solution Approach 2:
The patent replaces the manual mechanical process of annotation with an automated optical and computational system. Instead of opticians manually marking the corneal vertex position, the system uses cameras to capture images of the light pattern on the cornea, and computer algorithms to automatically determine the vertex location, substituting human manual work with automated image processing.
3Ease of operation
If reflex-based annotation is used, then detection is easier compared to pupil center, but many opticians still prefer pupil-based annotation
Solution Approach 1:
The system is designed to be universal by accommodating both reflex-based and pupil-based detection methods. The light pattern can be projected to create a reflex on the cornea, and the same imaging system can also capture the pupil. The software can process either type of annotation, making the system versatile and acceptable to opticians who prefer either method.
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
Enables more precise determination of corneal vertex position and centering parameters, facilitating accurate lens placement and grinding in eyeglass frames, even with non-overlapping camera views and varying recording directions.
Implementation Method 1
a reflex-based evaluation is to be carried out by generating a flash of light, preferably by means of an LED, when taking the images, whereby the position of the corneal vertex in space is determined to a first approximation as the position of the reflection point of the flash of light on the cornea
Data Source
Figure 1a
Figure 1b
Figure 2
AI summary
The invention relates to a computer-implemented method for detecting a corneal vertex, wherein a front image of a head and a side image of the head are provided, which are calibrated to each other, and wherein the position of the corneal vertex in space is determined from the front image and the side image by means of geometric position determination.