Fingerprint Enrollment Using Diversity-Based Selective Storage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fingerprint enrollment and recognition processes are inefficient and slow due to the small size of fingerprint sensing regions in portable devices, which limits the accuracy and speed of fingerprint enrollment and verification.
Innovation Solution
A processor-implemented method that performs matching between input fingerprint images and enrolled images, calculates the degree of diversity in overlapping regions, and determines whether to store the input fingerprint based on similarity scores and enrollment scores, allowing for selective enrollment and efficient fingerprint recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Volume of moving object
If the fingerprint sensing region size is decreased to fit portable devices, then device portability is improved, but fingerprint enrollment efficiency and recognition accuracy deteriorate
Solution Approach 1:
The patent divides the fingerprint matching process into multiple stages: initial quick matching using down-scaled images for fast comparison, followed by detailed matching using original resolution images. This segmentation allows the system to rapidly filter candidates and then perform accurate verification, maintaining high enrollment efficiency despite the small sensing region.
Solution Approach 2:
The patent implements a two-pass matching approach where only a portion of the fingerprint image (down-scaled version) is processed first for quick comparison, and then the full-resolution image is processed only for matched candidates. This partial processing strategy significantly reduces computational load and improves enrollment speed while maintaining accuracy.
2Volume of moving object
If the fingerprint sensing region size is decreased to fit portable devices, then device portability is improved, but fingerprint recognition accuracy deteriorates
Solution Approach 1:
The patent segments the recognition process into initial screening and final verification stages. The down-scaled image matching provides rapid initial screening, while the original resolution image matching ensures high accuracy for the final verification, thus maintaining recognition accuracy despite the small sensing region.
Solution Approach 2:
The patent performs preliminary matching using down-scaled images before conducting the final accurate matching with original resolution images. This preliminary action filters out non-matching fingerprints early, ensuring that only promising candidates undergo the computationally intensive high-precision matching, thereby maintaining accuracy while improving efficiency.
3Reliability
If multiple fingerprint images are stored to improve recognition reliability, then authentication reliability is improved, but device memory requirements and processing complexity increase
Solution Approach 1:
The patent processes only a down-scaled version of fingerprint images for initial matching, which reduces the computational complexity and memory requirements compared to processing full-resolution images. This partial processing approach allows the system to store and compare multiple fingerprint images while maintaining manageable device complexity.
Data Source
AI summary
A processor fingerprint enrollment method and apparatus is disclosed. A processor implemented fingerprint enrollment method includes performing a matching between a received input fingerprint image of a user and one or more enrolled fingerprint images, and selectively, based on a result of the matching identifying a matched enrolled fingerprint image from the one or more enrolled fingerprint image and based on a calculated degree of diversity in the fingerprint corresponding to an overlapping region between the input fingerprint image and the matched enrolled fingerprint image, storing the input fingerprint as another enrolled fingerprint image.