Eyelid Shape Estimation for Iris Segmentation Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing iris segmentation techniques face challenges in accurately estimating the portion of the iris occluded by eyelids, which affects the quality of biometric authentication and requires extensive calculations, as the entire iris is rarely visible due to occlusions by eyelids and eyelashes.
Innovation Solution
The method involves generating an eye-box over the eye image with radial lines to estimate the eyelid shape, using edge detection and iterative processes to fit curves to candidate points, excluding outliers, and refining the fit to improve the accuracy of eyelid shape estimation, thereby enhancing iris segmentation and reducing computational load.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing iris segmentation techniques are used to estimate occluded iris portions, then biometric authentication can be performed, but computational complexity increases and accuracy decreases due to extensive calculations required
Solution Approach 1:
The patent divides the eye image into distinct regions (iris, eyelid, eyelashes, sclera) and processes each segment separately. By segmenting the image and focusing calculations only on relevant regions rather than the entire image, the computational complexity is reduced while maintaining segmentation accuracy.
Solution Approach 2:
The patent performs preliminary detection of eyelid and eyelash positions before conducting detailed iris segmentation. This preliminary action identifies occluded regions in advance, allowing the system to exclude these areas from subsequent calculations, thereby reducing computational load while preserving accuracy in visible iris portions.
2Loss of information
If the entire eye image is processed for iris segmentation, then complete iris information can be obtained, but processing time and computational resources increase
Solution Approach 1:
The patent extracts and excludes occluded portions of the iris that are blocked by eyelids and eyelashes from the processing pipeline. By taking out these irrelevant or unusable segments, the system reduces processing time and resources while maintaining completeness of the usable iris data for authentication.
Solution Approach 2:
The patent applies partial action by processing only the visible, unoccluded portions of the iris rather than attempting to process the entire iris area. This approach accepts that some portions may be occluded and focuses computational efforts only on the partial visible regions, reducing processing time while obtaining sufficient information for authentication.
3Measurement precision
If conventional eyelid shape estimation methods are used, then eyelid boundaries can be detected, but measurement accuracy is insufficient for precise iris segmentation
Solution Approach 1:
The patent introduces an intermediary iterative fitting process that refines eyelid shape estimation by repeatedly adjusting the fitted curve to better match detected edge points. This intermediary refinement step improves measurement precision of eyelid boundaries without significantly impacting overall authentication efficiency, as the iterative process converges quickly.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Systems and methods for eyelid shape estimation are disclosed. In one aspect, after receiving an eye image of an eye (e.g., from an image capture device), an eye-box is generated over an iris of the eye in the eye image. A plurality of radial lines can be generated from approximately the center of the eye-box to an upper edge or a lower edge of the eye box. Candidate points can be determined to have local maximum derivatives along the plurality of radial lines. From the candidate points, an eyelid shape curve (e.g., for an upper eyelid or a lower eyelid) can be determined by fitting a curve (e.g., a parabola or a polynomial) to the candidate points or a subset of the candidate points.