Biometric Image Resizing Using Frequency-Based Ridge Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for resizing biometric images, such as fingerprints, lack the accuracy needed for reliable comparison and identification due to varying resolutions and occlusions, leading to inefficient and costly identification processes.
Innovation Solution
A method that determines a first frequency of biometric characteristics using Fourier transformation or density, adjusts it based on properties like aspect ratio, and resizes the image to a target resolution using a regression function, ensuring accurate resizing and comparison.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional resizing methods are used on biometric images, then the image can be resized to match reference resolution, but the resizing accuracy is insufficient for reliable biometric comparison
Solution Approach 1:
The patent changes the parameter used for resizing from fixed geometric dimensions to frequency-based measurements. By determining the frequency of biometric characteristics (ridges, lines, patterns) in both the input and reference images, the system calculates a scaling factor that preserves the true spatial relationships of biometric features, achieving accurate resizing for reliable comparison
Solution Approach 2:
The patent replaces conventional mechanical/geometric resizing methods with a frequency-domain approach. Instead of simply scaling pixels based on image dimensions, the system uses Fourier transformation or frequency analysis to understand the spatial periodicity of biometric characteristics, then applies intelligent scaling that maintains the integrity of these characteristics
2Reliability
If multiple resizing attempts are made to handle outliers, then identification accuracy improves, but processing cost and time increase proportionally
Solution Approach 1:
The patent performs preliminary frequency analysis and scaling factor calculation before the actual biometric comparison. By determining the appropriate scaling factor upfront based on frequency measurements, the system avoids the need for multiple retry attempts, handling outliers in a single pass and maintaining high processing efficiency
Solution Approach 2:
The system uses frequency measurement as feedback to determine the optimal scaling factor. By analyzing the frequency characteristics of the input image and comparing it with reference frequency data, the system automatically adjusts the scaling to achieve accurate matching without requiring manual intervention or multiple attempts
3Ease of operation
If fixed width assumption is used for resizing, then the process is simple and fast, but it fails when finger width varies or fingers are occluded
Solution Approach 1:
The patent transitions from a static, fixed-width resizing approach to a dynamic, adaptive method. The scaling factor is not predetermined but is dynamically calculated based on the measured frequency of biometric characteristics in the specific input image, allowing the system to adapt to varying finger widths, orientations, and occlusion conditions
Solution Approach 2:
The system changes the parameter basis for resizing from fixed geometric assumptions to measured frequency characteristics. By determining the actual frequency of ridges and biometric patterns in the input image, the system calculates an appropriate scaling factor that adapts to the specific conditions of each capture, including variations in finger width and occlusion
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
Enables highly accurate resizing of biometric images to any desired resolution, facilitating reliable user identification and reducing the need for multiple attempts, thus improving security and efficiency.
Implementation Method 1
determining a first frequency of at least part of the plurality of biometric characteristics in the image
Data Source
AI summary
Method for resizing an image, the method comprising the steps: obtaining an image of a user comprising at least one biometric identifier, wherein the biometric identifier comprises a plurality of biometric characteristics; determining a first frequency of at least part of the plurality of the biometric characteristics in the image; determining a second frequency based on the first frequency and a property of the biometric identifier; resizing the image based on a ratio of the first and the second frequency; obtaining a resized image.