Fingerprint Verification based on Gabor Filter Enhancement

TL;DR

This paper introduces a fingerprint verification method utilizing Gabor filter enhancement for improved minutiae extraction, combining global and local features to enhance ridge and valley details, resulting in better sensitivity and specificity.

Contribution

The paper presents a novel fingerprint verification algorithm that integrates Gabor filter enhancement with a new post-processing technique for minutiae extraction, improving accuracy over existing methods.

Findings

Higher sensitivity and specificity compared to previous algorithms

Effective extraction of ridge orientation and frequency

Improved minutiae detection accuracy

Abstract

Human fingerprints are reliable characteristics for personnel identification as it is unique and persistence. A fingerprint pattern consists of ridges, valleys and minutiae. In this paper we propose Fingerprint Verification based on Gabor Filter Enhancement (FVGFE) algorithm for minutiae feature extraction and post processing based on 9 pixel neighborhood. A global feature extraction and fingerprints enhancement are based on Hong enhancement method which is simultaneously able to extract local ridge orientation and ridge frequency. It is observed that the Sensitivity and Specificity values are better compared to the existing algorithms.