Iris Recognition Fraud Detection via Inter-Eye Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional iris recognition methods are unreliable when individuals wear cosmetic contact lenses, as these lenses can alter the iris pattern, making fraud detection difficult or impossible.
Innovation Solution
A method that captures images of both eyes, extracts and correlates iris characteristics, evaluates correlation coefficients, and signals fraud attempts by identifying artificial patterns, particularly through angular offset and radial direction analysis, with additional checks for iris diameter and comparison with stored fraudulent patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If cosmetic contact lenses are worn to alter iris appearance, then the ability to conceal identity is improved, but the reliability of iris recognition is worsened
Solution Approach 1:
The method segments the iris recognition process into multiple independent analysis stages: initial characteristic extraction, correlation evaluation within each eye, correlation evaluation between eyes, and fraud detection. This segmentation allows the system to identify artificial patterns at different levels without requiring complete pattern matching, thereby maintaining reliability even when cosmetic lenses are worn.
Solution Approach 2:
The patent introduces correlation coefficients as an intermediary metric between the raw iris characteristics and the final identification decision. By evaluating correlations of characteristics within each eye and between both eyes, the system creates an intermediate layer of analysis that can detect artificial patterns introduced by cosmetic lenses before making the final identification determination.
2Productivity
If conventional iris recognition methods are used, then the process is simple and quick, but fraud detection capability is lost when cosmetic contact lenses are worn
Solution Approach 1:
The method performs preliminary correlation evaluations of iris characteristics within each eye before comparing with stored identification data. This preliminary action of evaluating correlations within the first set of first characteristics and within the second set of second characteristics allows the system to quickly identify potential fraud cases before committing to full identification processing, maintaining efficiency while enhancing detection capability.
Solution Approach 2:
The patent applies partial action by selectively evaluating correlations only for characteristics that show suspicious patterns. Rather than analyzing all characteristics in full detail, the system first evaluates correlations to identify potential fraud indicators, then focuses detailed analysis only where needed, thereby maintaining processing speed while improving fraud detection.
3Measurement precision
If correlation evaluation of all characteristics is performed, then fraud detection accuracy is improved, but processing time increases
Solution Approach 1:
The method performs partial correlation evaluation by first assessing correlations within each eye separately, then only performing inter-eye correlation evaluation when necessary. This selective approach maintains high detection accuracy for obvious fraud cases while reducing processing time for legitimate users whose iris patterns show normal correlation characteristics.
Solution Approach 2:
The correlation evaluation process is segmented into distinct stages: intra-eye correlation evaluation for the first image, intra-eye correlation evaluation for the second image, and then inter-eye correlation evaluation. This segmentation allows the system to quickly eliminate legitimate cases at earlier stages while dedicating more computational resources to suspicious cases that require full analysis.
Data Source
AI summary
A method of detecting fraud during identification by iris recognition, the method comprising the following steps:capturing an image of each eye of a person for identification (50), namely a first image (61) and a second image (71);extracting a first set of first characteristics from the first image (61);extracting a second set of second characteristics from the second image (71);evaluating a correlation coefficient between the first and second characteristics; andas a function of the value of the correlation coefficient, signaling an attempt at fraud or continuing with identification by eye recognition.An identification terminal arranged to perform the method.