Biometric Image Resizing Using Ridge Frequency Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for resizing biometric images, such as fingerprints, fail to achieve accurate resizing across varying user biometric characteristics and capture conditions, leading to unreliable comparisons and increased costs in identification processes.
Innovation Solution
A method involving determining the first frequency of biometric characteristics, such as fingerprint ridges, through Fourier transformation or density analysis, and using a regression function to predict a second frequency based on user properties, allowing for precise resizing to a target resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional resizing methods are used on biometric images captured by cameras, then the images can be resized to match reference resolution, but the resizing accuracy is insufficient for reliable biometric comparison
Solution Approach 1:
The patent changes the approach from direct geometric resizing to frequency-domain analysis. By determining the first frequency of biometric characteristics in the captured image and comparing it with the second frequency (ridge frequency) of the reference image, the system calculates an accurate scaling factor that preserves the true dimensions of biometric features, thereby achieving both high resizing accuracy and reliable comparison results
Solution Approach 2:
The patent replaces conventional mechanical/geometric resizing methods with a frequency-based analytical approach. Instead of simply scaling pixels based on image dimensions, the system uses Fourier transformation to analyze the frequency characteristics of biometric features, determining the actual physical dimensions through frequency domain analysis, which provides superior accuracy for biometric comparison
2Reliability
If multiple resizing attempts are performed to handle outliers with atypical ridge frequencies, then identification accuracy improves, but processing costs increase proportionally
Solution Approach 1:
The patent performs preliminary frequency analysis on the captured biometric image before attempting comparison with reference images. By determining the first frequency and calculating the scaling factor in advance, the system prepares the image with accurate dimensions upfront, eliminating the need for multiple retry attempts and reducing processing costs while maintaining high identification accuracy
Solution Approach 2:
The patent implements a feedback mechanism where the determined first frequency from the captured image is compared with the known second frequency (ridge frequency) of biometric characteristics. This feedback loop allows the system to automatically calculate the correct scaling factor and adjust the image dimensions accordingly, handling outliers efficiently without requiring multiple manual attempts
3Ease of operation
If biometric images are captured at varying distances and resolutions by mobile cameras, then user convenience improves, but the resulting images require complex resizing to match reference resolution
Solution Approach 1:
The patent transforms the resizing problem from a geometric scaling task to a frequency analysis task. By extracting the first frequency from the captured image and comparing it with the reference ridge frequency, the system determines the actual scaling factor needed, simplifying the complexity while maintaining accuracy across varying capture conditions
Solution Approach 2:
The patent creates a universal resizing method that works across all capture conditions (varying distances, resolutions, and devices). The frequency-based approach is device-agnostic and can handle any captured image by analyzing its biometric characteristic frequencies, making the solution universally applicable without requiring device-specific calibration or complex condition-based processing
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, facilitating reliable user identification and reducing the need for multiple attempts, thus lowering costs and improving security access processes.
Implementation Method 1
determining a first frequency of at least part of the plurality of the biometric characteristics in the image
Data Source
Figure 1a
Figure 1b
Figure 2
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.