Fingerprint Authentication via Stitched Mosaic Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint authentication systems face challenges in ensuring accurate matches when the authentication image is not well aligned with enrolled images, especially in systems without hardware guidance, and require significant computational power and time.
Innovation Solution
The method involves creating a mosaic from partial enrolment images to form an enrolled image, determining an area of interest, and matching the authentication image with the mosaic, allowing for efficient authentication regardless of rotation or partial overlap, and reducing computational requirements by matching against a single image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple partial enrolment images are stored for matching, then authentication reliability is improved, but device complexity and computational requirements increase
Solution Approach 1:
The patent merges multiple partial enrolment images into a single composite enrolment image by stitching them together. This composite image contains all fingerprint features from multiple angles and positions, allowing the system to maintain high authentication reliability while simplifying the matching process to a single image comparison instead of comparing against multiple separate partial images
2Measurement precision
If multiple partial enrolment images are matched against authentication image, then matching accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary stitching of multiple partial enrolment images into a complete composite image during the enrolment phase. This preliminary action ensures all fingerprint features are captured and integrated before authentication occurs, so that during authentication only a single fast comparison is needed rather than multiple sequential comparisons against partial images
3Area of stationary object
If small fingerprint sensors are used, then device size is reduced, but area coverage and overlap between enrolment and authentication images become insufficient
Solution Approach 1:
The patent transitions from a single two-dimensional sensor view to a composite image that integrates information from multiple angles and positions. By stitching partial images taken at different orientations, the system creates a virtual enlarged fingerprint image that provides sufficient area coverage and feature overlap for reliable authentication while继续使用小型传感器
4Ease of operation
If authentication image is not well aligned with enrolled images, then ease of operation is improved, but matching accuracy deteriorates
Solution Approach 1:
The patent combines multiple partial enrolment images from different angles and positions into a single composite image that contains the complete fingerprint pattern. This composite image serves as a comprehensive reference that can match against authentication images regardless of their alignment or orientation, thereby maintaining high matching accuracy while allowing users to place their fingers in various positions without precise alignment requirements
Data Source
AI summary
The present invention concerns a method, a system, a device of authenticating a user using a body part including biometric information. The method comprises stitching partial enrolment images comprising different views of the user's body part into at least one mosaic and thereby creating an enrolled image, acquiring an authentication image of the body part, wherein the size of the authentication image is a fraction of the enrolled image, determining at least one area of interest in the enrolled image, and matching the authentication image with image data of the determined area of interest in the enrolled image.