Front-Image Eye Protrusion Estimation Using Depth Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for estimating eye protrusion value using front facial images are inaccurate and require specialized equipment, making it difficult for individuals to conveniently and accurately measure this value outside a medical setting.
Innovation Solution
A method utilizing a personal electronic device to capture a front facial image, generate a depth image, and apply a pre-trained estimation model that combines facial and depth images to accurately estimate eye protrusion value.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a side image is used to estimate eye protrusion value, then measurement accuracy is improved, but ease of operation deteriorates because the user must face forward and capture the side of the eye from a perpendicular direction
Solution Approach 1:
The patent inverts the conventional approach by using a front image instead of a side image for eye protrusion estimation. This reversal allows users to easily capture images by simply facing the camera while maintaining measurement accuracy through specialized processing of the front image to extract depth information about eyeball protrusion.
Solution Approach 2:
The patent transitions from 2D front image analysis to 3D depth information extraction. By generating a depth map from the front image and combining it with 3D facial landmark detection, the system recovers depth dimensions (z-axis differences) that are not visible in the original 2D image, enabling accurate protrusion measurement without requiring side-view imaging.
2Ease of operation
If a 3D facial landmark detection model is used to compute eye protrusion from front image, then ease of operation is improved, but measurement precision deteriorates because accurate eye protrusion value cannot be computed using only the 3D facial landmark detection model
Solution Approach 1:
The patent merges multiple processing components to achieve accurate eye protrusion estimation from front images. It combines 3D facial landmark detection results with depth map generation and specialized protrusion calculation algorithms, integrating these elements to overcome the limitations of any single method and achieve both ease of operation and measurement precision.
Solution Approach 2:
The patent introduces a depth map as an intermediary element between the front image and the final protrusion measurement. The depth map serves as a mediator that translates 2D image information into depth-related data, enabling the system to estimate eyeball protrusion accurately without directly capturing side-view geometry.
3Ease of operation
If a depth estimation model is used to generate depth map from front image, then ease of operation is improved, but measurement precision deteriorates because accurate eye protrusion value cannot be computed using only the depth map
Solution Approach 1:
The patent combines depth map information with 3D facial landmark detection results to achieve accurate eye protrusion estimation. By merging these two data sources, the system overcomes the limitation of using only the depth map, as the combination provides both depth context and precise anatomical reference points needed for accurate protrusion measurement.
Data Source
AI summary
A method for estimating an eye protrusion value, the method including: obtaining an image representing at least one eye of the subject, wherein the image includes a plurality of pixels assigned a value corresponding to at least one of brightness and color; obtaining a pre-processed image by performing a previously stored pre-processing for the image; obtaining a depth image corresponding to the pre-processed image by applying the pre-processed image to a pre-trained depth image generation model, wherein the depth image includes a plurality of pixels, wherein each of the plurality of pixels of the depth image is assigned a depth value representing a relative distance of an object corresponding to each of a plurality of pixels of the pre-processed image; and estimating an eye protrusion value for the eye of the subject by applying both the pre-processed image and the depth image to a pre-trained eye protrusion value estimation model.