Biometric Authentication Device Using Weighted Feature Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Biometrics authentication systems face a high false acceptance rate when features representing palm patterns are included, as these features vary less than vein features, and methods that physically separate palm patterns are not applicable, leading to increased false positives.
Innovation Solution
A biometrics authentication device and method that utilize a combination of non-directional and directional feature extraction processes, including Gabor filters and binarization units, to generate and match features from palm region images, allowing for robust authentication without relying on physical separation of palm patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If biometrics information includes features representing palm pattern, then the authentication system can operate without physical separation methods, but the false acceptance rate increases
Solution Approach 1:
The patent segments biometric features into two distinct groups: non-directional features (including palm patterns) and directional features (vein patterns). By processing these feature groups separately through different extraction and matching pipelines, the system can utilize both feature types while controlling their individual contributions to authentication decisions, thereby maintaining operational simplicity without physical separation while managing the false acceptance rate through independent feature evaluation
Solution Approach 2:
The patent changes the evaluation parameters by introducing separate scoring mechanisms for non-directional and directional features. By adjusting the weightings and threshold parameters for each feature group independently, the system can optimize the balance between utilizing palm pattern information for ease of operation and controlling the false acceptance rate through parameter tuning in the feature matching and score determination processes
2Productivity
If features representing palm pattern are extracted from image, then the authentication can proceed without physical separation, but the false acceptance rate increases especially when melanin is heavily deposited
Solution Approach 1:
The patent segments the authentication process into parallel streams: one processing non-directional features (palm patterns) for quick initial assessment and another processing directional features (veins) for higher reliability verification. This segmentation allows the system to maintain high productivity through rapid non-directional feature matching while using directional features to suppress false acceptances, particularly in cases with heavy melanin deposition where vein patterns remain distinguishable
Solution Approach 2:
The patent introduces an intermediary scoring and weight determination mechanism that evaluates both non-directional and directional feature matches. This intermediary process acts as a mediator between the fast but less reliable non-directional feature matching and the more reliable but slower directional feature matching, combining their results with appropriate weightings to achieve both high productivity and low false acceptance rates
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution effectively suppresses the increase in false acceptance rates by actively and passively using non-directional and directional feature groups in the authentication process, improving the accuracy and robustness of biometric authentication.
Implementation Method 1
a non-directional feature group generation process that generates a non-directional feature group from the palm region image, and a directional feature group generation process that generates a directional feature group from the palm region image
Implementation Method 2
binarization units, to generate and match features from palm region images
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A biometrics authentication device 1 is configured to include a non-directional feature generation process unit 8 configured to extract, from an image, directional features corresponding to directions different from each other and to generate a non-directional feature on the basis of the directional features; a directional feature generation process unit 9 configured to extract, from an image, directional features corresponding to directions different from each other and to select, from among the directional features, a reference directional feature corresponding to a reference direction; a non-directional feature matching process unit 10 configured to obtain a first degree of similarity between the non-directional feature and a registered non-directional feature stored in a storage unit 7; a directional feature matching process unit 11 configured to obtain a second degree of similarity between the reference directional feature and a registered reference directional feature stored in the storage unit 7; and a determination unit 6 configured to make a weight of the second degree of similarity smaller than a weight of the first degree of similarity and to determine whether or not a subject is a person to be authenticated, by using the first degree of similarity and the second degree of similarity.