Fingerprint Image Noise Suppression via Frequency Domain Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing fingerprint recognition technologies face challenges in effectively suppressing noise components in fingerprint images, particularly low-frequency noise due to measurement errors and high-frequency noise caused by device artifacts, which can lead to inaccurate authentication results.
Innovation Solution
A processor-implemented method that transforms a fingerprint image into a frequency domain, suppresses low-frequency noise by masking it with a statistically representative value, and optionally suppresses high-frequency noise by masking it with a median power value, thereby restoring a noise-reduced fingerprint image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If fingerprint images are processed using traditional noise filtering methods, then processing speed is maintained, but noise components (low-frequency measurement errors and high-frequency device artifacts) cannot be effectively suppressed, leading to reduced authentication accuracy
Solution Approach 1:
The patent segments the frequency spectrum into distinct regions (low-frequency region for measurement errors, mid-frequency region for fingerprint features, and high-frequency region for device artifacts). By applying different processing strategies to each segment, the method achieves effective noise suppression while preserving authentication features. The frequency domain segmentation allows targeted suppression of specific noise components without affecting the entire spectrum uniformly.
Solution Approach 2:
The patent transforms the fingerprint image from the spatial domain to the frequency domain using Fourier transform. This dimensional transformation enables the system to visualize and process different noise components (low-frequency and high-frequency) as distinct regions in the frequency spectrum, making it possible to apply selective suppression strategies that would be difficult to implement in the spatial domain.
2Reliability
If frequency domain transformation and noise suppression is applied, then noise components are effectively removed, but processing time and computational load increase
Solution Approach 1:
The patent applies partial action by selectively suppressing only specific frequency regions containing noise components rather than processing the entire frequency spectrum uniformly. The low-frequency region suppression targets measurement errors, while high-frequency region suppression targets device artifacts. This selective approach reduces unnecessary computational effort compared to comprehensive frequency domain processing, thereby lowering processing time while maintaining authentication reliability.
3Measurement precision
If aggressive noise suppression is applied to remove all noise components, then authentication accuracy improves, but fingerprint features may be distorted or lost
Solution Approach 1:
The patent applies local quality by using different suppression strategies for different frequency regions. The low-frequency region uses one suppression approach tailored for measurement errors, while the high-frequency region uses another approach tailored for device artifacts. The mid-frequency region containing fingerprint features is preserved with minimal suppression. This region-specific quality control ensures effective noise removal while maintaining fingerprint feature integrity.
Data Source
AI summary
A processor-implemented method with image preprocessing includes: transforming a fingerprint image into a frequency domain; suppressing a low-frequency region corresponding to a first noise component of the fingerprint image in the frequency domain; and restoring the frequency domain, in which the low-frequency region is suppressed, as an image.