Transformed Biometric Matching With Error-Correcting Code Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric systems face challenges in maintaining confidentiality and reliability due to substantial spatial deformations caused by irreversible transformation functions, leading to spurious comparison results and limited confidentiality protection.
Innovation Solution
The method involves obtaining and comparing transformed biometric data sets using an irreversible transformation function, with the comparison performed at the bit level based on an error-correcting code word, allowing for dependable results even with substantial deformations and enhanced confidentiality protection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If an irreversible transformation function is applied to biometric data to protect confidentiality, then the level of confidentiality protection increases, but the comparison algorithms yield spurious results due to substantial spatial deformations
Solution Approach 1:
The patent segments the transformed biometric data into multiple blocks and processes each block independently through the comparison algorithm. This segmentation allows the system to handle spatial deformations locally rather than globally, maintaining comparison reliability even when substantial transformations are applied to the overall data structure for confidentiality protection.
Solution Approach 2:
The patent introduces an intermediary comparison process that operates on transformed data blocks rather than directly on original biometric data. This intermediary approach uses reference blocks stored in a database to mediate the comparison, enabling reliable authentication decisions to be made from transformed data without revealing original biometric information.
2Object-affected harmful factors
If the transformation function is made more substantial to increase confidentiality, then the level of protection increases, but the comparison stage produces more spurious results
Solution Approach 1:
By dividing transformed biometric data into multiple manageable blocks and comparing them independently, the system can tolerate greater overall transformation while maintaining local comparison accuracy. Each block's comparison is less sensitive to global spatial deformations, allowing more substantial transformations for enhanced confidentiality.
Solution Approach 2:
The patent changes the comparison parameters from global spatial metrics to local block-based metrics. This parameter transformation allows the system to accept more substantial overall transformations while maintaining measurement precision at the block level, thereby enabling higher confidentiality protection without sacrificing comparison accuracy.
Data Source
AI summary
Biometric data relating to a biological part are processed by obtaining, on the one hand, a first set of transformed biometric data (f(B1)) by applying at least one irreversible transformation to a first set of biometric data (B1), and, on the other hand, a second set of transformed biometric data (f(B2)) by applying said transformation to a second set of biometric data (B2). Thereafter, a decision is made as to whether the second biometric data set corresponds to the first biometric data set on the basis of a comparison between the first transformed biometric data set and the second transformed biometric data set, said comparison being performed at the bit level of a digital representation of said first and second transformed biometric data sets as a function of an error corrector code word.