Fake Fingerprint Detection Using Registered Image Comparison
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint recognition technologies face challenges in accurately identifying fake fingerprints, particularly due to high costs associated with additional optical sensors and the risk of duplication using features like spectrum, reflectance, and sweat pore density.
Innovation Solution
A fake fingerprint recognition device and method that calculates a fingerprint index by comparing input images with registered images stored in advance, using a processor to determine if the index exceeds a threshold, thereby identifying fake fingerprints without additional optical sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If live fingerprint recognition using additional optical sensors is implemented, then fake fingerprint detection capability is improved, but device cost increases
Solution Approach 1:
The patent extracts the detection function from separate optical sensors and integrates it into the existing fingerprint sensor by analyzing image data features (edge information, gradient, curvature) that naturally exist in fingerprint images. This eliminates the need for additional sensors while maintaining detection capability.
Solution Approach 2:
The patent uses the existing fingerprint image data to create derived features (edge maps, gradient information, curvature values) that serve as indicators for fake fingerprint detection, copying the detection function from specialized sensors to general image processing.
2Reliability
If fingerprint recognition based on features such as spectrum, reflectance, and sweat pore density is used, then fake fingerprint identification capability is improved, but vulnerability to duplication attacks increases
Solution Approach 1:
The patent focuses on local geometric properties of fingerprint features (edge information, gradient, curvature) rather than global biometric characteristics. These local structural properties are harder to duplicate while still being present in both real and fake fingerprints, creating a more robust detection mechanism.
Solution Approach 2:
Instead of trying to detect what makes fingerprints authentic (biological features like sweat pores), the patent inverts the approach by detecting structural inconsistencies in the geometric properties of fingerprint patterns, which reveal fake fingerprints through their imperfect replication of these properties.
3Measurement precision
If fingerprint images are compared using registered images stored in advance, then accuracy in identifying fake fingerprints is improved, but data storage requirements increase
Solution Approach 1:
The patent performs preliminary processing of fingerprint images during registration to extract and store only the essential geometric features (edge information, gradient, curvature) rather than storing complete high-resolution images. This reduces storage requirements while maintaining comparison accuracy.
Solution Approach 2:
The patent extracts only the necessary geometric features from complete fingerprint images for storage and comparison purposes, separating the essential detection information from the full image data to reduce storage requirements while maintaining identification accuracy.
Data Source
AI summary
A fake fingerprint recognition device includes a memory and a processor. The memory is configured to store at least one command. The processor is configured to read the at least one command to execute following steps: receiving an input fingerprint image; calculating a fingerprint index according to the input fingerprint image and a registered fingerprint image in a fingerprint database, wherein the input fingerprint image corresponds to the registered fingerprint image, and the registered fingerprint image is stored in the fingerprint database in advance; determining whether the fingerprint index is larger than a predetermined index threshold; and if the fingerprint index is larger than the predetermined index threshold, determining that the input fingerprint image is a fake fingerprint image.