FDeblur-GAN: Fingerprint Deblurring using Generative Adversarial Network

TL;DR

This paper introduces FDeblur-GAN, a novel deep learning model based on cGANs for deblurring fingerprint images, incorporating auxiliary networks to preserve ridge details and identity information, significantly improving matching accuracy.

Contribution

The paper presents a multi-stage cGAN-based fingerprint deblurring model with auxiliary ridge extraction and verification networks, enhancing detail preservation and identity retention compared to prior methods.

Findings

Achieved 95.18% matching accuracy on blurred fingerprint dataset.

Successfully integrated ridge and identity verification sub-networks.

Improved deblurring quality enhances automated fingerprint recognition.

Abstract

While working with fingerprint images acquired from crime scenes, mobile cameras, or low-quality sensors, it becomes difficult for automated identification systems to verify the identity due to image blur and distortion. We propose a fingerprint deblurring model FDeblur-GAN, based on the conditional Generative Adversarial Networks (cGANs) and multi-stage framework of the stack GAN. Additionally, we integrate two auxiliary sub-networks into the model for the deblurring task. The first sub-network is a ridge extractor model. It is added to generate ridge maps to ensure that fingerprint information and minutiae are preserved in the deblurring process and prevent the model from generating erroneous minutiae. The second sub-network is a verifier that helps the generator to preserve the ID information during the generation process. Using a database of blurred fingerprints and corresponding ridge maps, the deep network learns to deblur from the input blurry samples. We evaluate the proposed method in combination with two different fingerprint matching algorithms. We achieved an accuracy of 95.18% on our fingerprint database for the task of matching deblurred and ground truth fingerprints.