Biometric Authentication Aligning Encrypted Feature Points
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional biological information authentication systems face challenges in protecting sensitive data from leakage and maintaining authentication accuracy, particularly when encryption is used, as decryption is required for collation, leading to potential data exposure and accuracy degradation.
Innovation Solution
The system separates position alignment and collation distance calculation processes, performing position alignment before transformation and executing these in distinct computer environments to reduce data exposure, and uses encrypted position correction information to align feature points without decrypting sensitive data, thereby enhancing security and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If encryption is applied to protect biological information, then security is improved, but authentication accuracy deteriorates due to required decryption for collation
Solution Approach 1:
The patent divides the authentication process into two independent stages: position alignment stage (where encrypted position correction information is decrypted only for alignment purposes) and collation stage (where encrypted feature information is collated directly without decryption). This segmentation allows security measures to be applied during collation while maintaining accuracy through position alignment, resolving the contradiction between security and accuracy.
Solution Approach 2:
The patent performs position alignment using encrypted position correction information before the collation process. By pre-aligning the feature information positions, the system prepares the data structure to maintain accuracy during subsequent encrypted collation, eliminating the need to decrypt sensitive biological information for the collation operation itself.
2Measurement precision
If position alignment is performed after transformation, then authentication accuracy is maintained, but data exposure risk increases due to decryption requirement
Solution Approach 1:
The patent inverts the conventional sequence by performing position alignment before transformation and collation, rather than after. The position correction information is applied to align feature points in the encrypted domain, and only then is the collation performed. This reversal eliminates the need to decrypt data for alignment operations, reducing data exposure risk while maintaining accuracy.
Solution Approach 2:
The patent introduces encrypted position correction information as an intermediary that enables position alignment without requiring decryption of the main biological information. This intermediary allows the system to perform necessary alignment operations in the encrypted domain, acting as a bridge between security requirements and accuracy needs.
3Reliability
If feature information is collated in encrypted form, then security is improved, but collation accuracy deteriorates due to encryption overhead
Solution Approach 1:
The patent segments the authentication process into position alignment (performed before collation using encrypted position correction information) and collation (performed in encrypted form). By separating these functions, the system can optimize each: position alignment maintains accuracy through pre-transformation alignment, while collation benefits from security through encrypted processing, resolving the contradiction between security and accuracy.
Data Source
AI summary
An obtaining unit of a biological information obtaining apparatus obtains biological information for authentication, and an extracting unit extracts first feature information from biological information for authentication. A generating unit of a biological information collating apparatus generates encrypted position correction information, and encrypted position correction information is transmitted to the biological information obtaining apparatus. A correcting unit decrypts encrypted position correction information to obtain position correction information, and by correcting first feature information, performs alignment between first feature information and second feature information extracted from biological information for registration. A transforming unit transforms corrected first feature information, and transmits transformed first feature information to the biological information collating apparatus. A collating unit collates transformed first feature information and transformed second feature information stored in a storing unit, and transmits a collation result to biological information obtaining apparatus.