Related Experiment Video
Updated: Jul 26, 2025
13:44
Detection of Architectural Distortion in Prior Mammograms via Analysis of Oriented Patterns
Published on: August 30, 2013
42.9K
An effective fingerprint orientation field estimation method using differential values of grayscale intensity
Ting-Wei Shen1, Mao-Hsiu Hsu2, Chun-Hsu Shen3
1Department of Mechanical Engineering, National Taiwan University, Taipei, Taiwan.
Peerj. Computer Science
|June 22, 2023
Summary
A new algorithm estimates fingerprint orientation fields (OF) using grayscale intensity differences. This method proves more reliable than existing techniques for low-quality and noisy fingerprints, improving accuracy in image processing.
Area of Science:
- Biometrics
- Image Processing
- Computer Vision
Background:
- Accurate fingerprint orientation field (OF) estimation is crucial for subsequent image processing tasks.
- Existing methods like gradient-based and power spectral density (PSD)-based approaches face challenges with low-quality or noisy fingerprint images.
- The accuracy of fingerprint enhancement techniques, such as Gabor filters, is directly influenced by the quality of OF estimation.
Purpose of the Study:
- To introduce and evaluate a novel orientation field (OF) estimation algorithm.
- To assess the accuracy and reliability of the proposed algorithm, particularly on degraded fingerprint images.
- To compare the performance of the new algorithm against established gradient-based and PSD-based methods.
Main Methods:
- Development of an OF estimation algorithm utilizing differential values of grayscale intensity.
- Experimental validation using fingerprint images subjected to Gaussian blurring and Gaussian white noise.
- Comparative analysis of the proposed method against gradient-based and PSD-based OF estimation techniques.
Main Results:
- The proposed OF estimation algorithm demonstrates superior reliability compared to gradient-based and PSD-based methods for low-quality fingerprints.
- In noisy fingerprint images, the proposed algorithm achieved reliability improvements of 6.46% over the gradient-based method.
- The algorithm showed a significant 32.93% increase in OF estimation reliability compared to the PSD-based method on noisy images.
Conclusions:
- The novel OF estimation algorithm based on grayscale intensity differences offers enhanced reliability, especially for challenging fingerprint images.
- This algorithm provides a valuable improvement for biometric systems dealing with noisy or low-quality fingerprint data.
- The findings suggest the proposed method is a robust alternative for critical fingerprint image processing applications.
