Biometric Feature Verification with Distributed Sample Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Biometric feature recognition systems are vulnerable to attacks such as forging, replay attacks, database hacking, transfer attacks, and matcher tampering, compromising their security and reliability in identity authentication processes.
Innovation Solution
A biometric feature verification method involving the extraction of a first biometric feature, sending a second biometric feature to a second end, and receiving a verification result by combining a first and second part of a sample feature stored in different ends, with the first part stored in a first end or a third end, and optionally processing the feature using a non-reversible algorithm to enhance security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If biometric features are stored in a centralized database for verification, then verification efficiency is improved, but security is worsened due to vulnerability to database hacking and centralized attacks
Solution Approach 1:
The patent segments the biometric feature data into multiple parts and distributes them across different storage locations (first storage location, second storage location, and third storage location). This segmentation prevents a single point of failure and requires attackers to compromise multiple locations simultaneously, thereby maintaining security while enabling efficient verification through coordinated access to segmented parts.
2Reliability
If biometric features are encrypted using non-reversible algorithms to prevent forgery, then security is improved, but verification capability is worsened because the original features cannot be recovered for comparison
Solution Approach 1:
The patent applies preliminary action by pre-processing the biometric features through non-reversible encryption before storage, and simultaneously pre-storing the encrypted versions in multiple locations. During verification, the same encryption process is applied to the input feature, and the encrypted versions are compared. This preliminary encryption action ensures security while maintaining verification capability through consistent transformation.
Solution Approach 2:
The patent creates encrypted copies of the biometric features and stores them in multiple locations. Instead of storing and comparing original features, the system stores and compares encrypted copies. This copying approach with non-reversible encryption maintains security (original cannot be recovered) while enabling verification (encrypted copies can be compared for matching).
Data Source
AI summary
The biometric feature verification method includes: extracting a first biometric feature from collected data; sending a second biometric feature to a second end according to the first biometric feature, the second biometric feature being the first biometric feature or being determined according to the first biometric feature; and receiving a result of verification performed by the second end on the second biometric feature by combining a first part of a sample feature and a second part of the sample feature stored in the second end, the first part being stored in the first end or a third end.