Edge-guided human eye image analysis for pupil-iris boundary accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing human eye image analyzing methods face challenges in accurately distinguishing the sclera and iris boundaries, especially under varying illumination, and often misidentify features like the pupil and iris areas due to rough segmentation maps and structural integrity issues.
Innovation Solution
An edge-guided human eye image analyzing method that uses a camera to collect images, processes them through a pre-trained contour generation network to obtain a detection contour map, and then uses a pre-trained edge-guided analyzing network for semantic segmentation and iterative parameter fitting to achieve accurate pupil-iris area division and parameter extraction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional semantic segmentation methods are used to divide pupil-iris areas, then the processing speed is relatively fast, but the boundary accuracy between sclera and iris deteriorates due to indistinct dividing lines and rough segmentation maps
Solution Approach 1:
The patent divides the image analysis into multiple stages: initial semantic segmentation to obtain rough pupil-iris area division, followed by edge detection to extract contour information, and finally iterative optimization to refine boundary accuracy. This multi-stage segmentation approach resolves the contradiction by breaking down the complex task into manageable steps that progressively improve boundary precision without overwhelming system complexity
Solution Approach 2:
The patent performs preliminary semantic segmentation to obtain an initial pupil-iris area division map before conducting edge detection and iterative optimization. This preliminary action provides a starting point that guides subsequent processing steps, enabling the system to achieve high boundary accuracy by building upon the initial segmentation results through progressive refinement
2Measurement precision
If deep learning-based methods are used to improve segmentation accuracy, then the boundary definition improves, but the structural integrity of the elliptical pupil-iris structure deteriorates due to missing or redundant parts
Solution Approach 1:
The patent implements an iterative optimization process where the system repeatedly refines the pupil-iris area division by comparing segmentation results with edge contour information and adjusting parameters accordingly. This feedback mechanism ensures that the final segmentation maintains both high accuracy and structural integrity by continuously correcting deviations from the expected elliptical structure
Solution Approach 2:
The patent employs iterative parameter optimization to adjust ellipse parameters (center position, semi-major axis, semi-minor axis, rotation angle) based on edge contour information. By dynamically changing these parameters through iterative refinement, the system maintains the structural integrity of the elliptical pupil-iris structure while achieving accurate segmentation boundaries
3Productivity
If simple segmentation approaches are used, then the computational efficiency is high, but the reliability of pupil-iris area division deteriorates under varying illumination and with distractors present
Solution Approach 1:
The patent segments the analysis process into distinct phases: initial semantic segmentation for efficiency, followed by edge detection for reliability, and iterative optimization for final accuracy. This segmentation allows the system to maintain computational efficiency while progressively improving reliability through each stage
Solution Approach 2:
The patent introduces edge contour information as an intermediary element that bridges the gap between initial semantic segmentation and final accurate division. This intermediary provides additional structural constraints that guide the optimization process, enabling the system to reliably distinguish pupil-iris areas from distractors and maintain division reliability under varying illumination conditions
Data Source
AI summary
The embodiments of the present disclosure disclose an edge-guided human eye image analyzing method. A specific implementation of this method comprises: collect a human eye image as an image to be detected; obtain a human eye detection contour map; obtain a semantic segmentation detection map and an initial human eye image detection fitting parameter; performing an iterative search on the initial human eye image detection fitting parameter to determine a target human eye image detection fitting parameter; sending the semantic segmentation detection map and the target human eye image detection fitting parameter as image analyzing results to a display terminal for display. This implementation improves the accuracy at the boundary dividing the pupil-iris area, and increases the structural integrity of the ellipse resulted from dividing the pupil-iris area. In addition, the iterative search can achieve a more accurate ellipse parameter fitting result.