Biometric Verification Axis Selection for Neighbor Discovery
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional verification processes face challenges in reducing arithmetic operation costs when feature points in biometric authentication systems, such as palm vein or fingerprint authentication, have a uniform distribution across axes, making it difficult to select an optimal datum axis for neighbor discovery.
Innovation Solution
The proposed solution involves a method to select an optimum datum axis by simulating neighbor discovery on each axis in an m-dimensional space, counting the comparison time number, and choosing the axis with the smallest number for feature points, thereby optimizing the verification process.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If feature points are verified in m-dimensional space using conventional methods, then verification accuracy is maintained, but arithmetic operation cost increases
Solution Approach 1:
The patent projects feature points from m-dimensional space onto a one-dimensional datum axis, transforming the verification problem from multi-dimensional distance calculation to one-dimensional coordinate comparison. This dimensionality reduction significantly decreases arithmetic operations while preserving verification accuracy through careful selection of the datum axis that maximizes feature point distribution spread.
Solution Approach 2:
The patent changes the parameter representation by selecting an optimal datum axis based on distribution width criteria. Instead of using fixed coordinate axes, the system dynamically determines which axis provides the greatest spread of projected coordinates, thereby optimizing the verification process for the specific feature point distribution while reducing computational complexity.
2Device complexity
If feature points are mapped to one-dimensional space for verification, then arithmetic operation cost is reduced, but verification accuracy may deteriorate
Solution Approach 1:
The patent dynamically selects the datum axis parameter based on the actual distribution characteristics of feature points. By calculating distribution width for each candidate axis and selecting the one with maximum spread, the system adapts the one-dimensional projection to preserve as much discriminative information as possible, thereby maintaining verification accuracy while achieving computational simplification.
3Productivity
If a datum coordinate axis is selected based on maximum distribution width, then neighbor discovery efficiency is improved, but difficulty in detecting and measuring increases
Solution Approach 1:
The patent performs preliminary calculation of distribution width for each candidate datum axis before the actual verification process. By pre-computing which axis provides maximum spread of feature point projections, the system prepares the optimal one-dimensional mapping in advance, ensuring efficient neighbor discovery during runtime without sacrificing accuracy.
Solution Approach 2:
The system automatically determines the optimal datum axis by analyzing the inherent distribution characteristics of the feature points themselves. The feature point distribution pattern guides the selection process, with the axis providing maximum spread naturally emerging from the data without requiring external intervention or complex heuristic rules.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method, for a verification process that performs neighbor discovery for one or more feature points projected to an m-dimensional space (m is a natural number equal to or greater than 2), includes: acquiring a feature point group including one or more feature points projected to coordinate values of the m-dimensional space ordered in a coordinate value order on each of two or more coordinate axes that define the m-dimensional space (m is a natural number equal to or greater than 2); selecting a datum axis on which a comparison time number in neighbor discovery is small, the comparison time number being obtained by performing simulation of neighbor discovery.