Fingerprint Authentication Calibration Using Multi-Order Functions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Fingerprint authentication devices face challenges in improving the characteristic distribution of sensor pixels and minimizing memory storage capacity for calibration data, leading to potential distortions in fingerprint images and increased storage needs.
Innovation Solution
A fingerprint authentication device and calibration method that include a biometric sensor, image processor, and authenticator, which utilize calibration data to calibrate sensing signals and generate fingerprint images, with the calibration data comprising multi-order functions and coefficients to minimize storage and correct pixel distribution issues.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If detailed calibration data is stored for each sensor pixel to correct characteristic distribution, then fingerprint image quality is improved, but memory storage capacity increases
Solution Approach 1:
The calibration data is segmented into two parts: (1) pixel-level correction values stored in a lookup table for each sensor pixel, and (2) region-level characteristic distribution parameters stored in a separate data structure. This segmentation allows the system to use compact pixel-level corrections while storing only essential region-level characteristics, reducing overall memory requirements while maintaining image quality.
Solution Approach 2:
The patent transforms the calibration approach from storing complete correction maps to storing calibration coefficients and characteristic distribution parameters in a multi-dimensional space. By representing calibration data as mathematical functions with coefficients rather than raw correction values for each pixel, the system achieves compression while preserving correction accuracy.
2Quantity of substance
If calibration data is minimized to reduce storage capacity, then memory usage is reduced, but characteristic distribution correction becomes insufficient
Solution Approach 1:
The system performs preliminary calibration during manufacturing to extract characteristic distribution parameters and generate calibration coefficients, which are then stored in compact form. This preliminary action captures the essential correction information in a compressed representation, allowing accurate correction during operation without storing extensive calibration data.
Solution Approach 2:
The patent changes the representation parameters of calibration data from pixel-by-pixel correction values to mathematical model parameters (calibration coefficients and characteristic distribution parameters). This parameter transformation enables the system to store minimal data while maintaining the ability to perform accurate corrections through mathematical computation.
Data Source
AI summary
Provided is a fingerprint authentication device comprising a biometric sensor that senses biometric information to generate a sensing signal; an image processor that extracts a calibration function based on calibration data and generates a fingerprint image based on the calibration function and the sensing signal; and an authenticator that performs fingerprint authentication by comparing the fingerprint image and registered fingerprint information. The calibration data includes information on a calibration coefficient of the calibration function.