Biometric Authentication Device Using Mutual Feature Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In large-scale biometric authentication systems, the 1:N authentication mode leads to increased processing time due to the need to match input biometric information with a large number of registered users, resulting in server load and latency issues, especially when using multiple pieces of biometric information.
Innovation Solution
A biometric authentication device that extracts mutual biometric information feature amounts representing the similarity between multiple pieces of biometric information, reducing the number of registered users to be matched by selecting based on similarity scores, thereby decreasing the operation amount and processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 1:N authentication mode is used to authenticate users against all registered users, then authentication coverage and security are improved, but processing time and server load increase significantly
Solution Approach 1:
The system performs preliminary extraction of mutual biometric information feature amounts between all pairs of biometric information pieces during registration. These pre-computed similarity scores are stored and reused during authentication, eliminating the need to re-evaluate all biometric pairs against all registered users, thus significantly reducing authentication time while maintaining security
Solution Approach 2:
The authentication process is segmented into two stages: (1) filtering stage using mutual biometric information feature amounts to identify candidate users, and (2) detailed matching stage for selected candidates. This segmentation reduces the search space from all registered users to only relevant candidates, decreasing processing time while preserving authentication accuracy
2Measurement precision
If multiple pieces of biometric information are used for authentication, then authentication accuracy is improved, but the number of comparisons and processing complexity increase
Solution Approach 1:
The system merges multiple biometric information pieces (e.g., fingerprints from different fingers) into a unified authentication decision by computing mutual feature amounts between all pairs. This combining approach allows the system to leverage the complementary information from multiple biometric sources to improve accuracy while managing complexity through efficient feature extraction and comparison
Data Source
AI summary
A biometric authentication device includes: a matching data generating unit which generates matching data from at least one biometric image representing a plurality of pieces of biometric information of a user; a mutual biometric information feature amount extraction unit which extracts a mutual biometric information feature amount representing the degree of similarity between two pieces of biometric information of the plurality of pieces of biometric information; a selection unit, which, for each of the first predetermined number of the registered users, computes a first selecting score based on the mutual biometric information feature amounts of the user and the registered users and selects the second predetermined number of registered users in order of decreasing the level of similarity based on the first selecting score of each registered users; and a matching unit which matches the matching data of the selected registered users with the matching data of the user.