Fingerprint Matching Using Pores Extracted with Machine Learning

• Ciro Moraes Rocha IFB
• Raimundo Claudio da Silva Vasconcelos IFB

Resumo

Fingerprint recognition is widely used for biometric applications due to its uniqueness and convenience. However, traditional systems often overlook level 3 features, particularly pores, which can improve precision, especially in partial or latent fingerprints. The integration of these features is highly beneficial in contexts such as forensic analysis, where complete fingerprint patterns are not always available. This work proposes a pore detection model, which is a convolution neural network with U-Net architecture followed by a pore matching algorithm, applied to the dataset L3-SF, a collection of extremely realistic, labeled synthetic fingerprints. The results demonstrated high effectiveness, achieving high accuracy in pore detection and matching even in partial fingerprint images, outperforming many existing methods and, consequently, the power of deep learning applied to biometrics.

Referências

D. Maltoni, D. Maio, A. K. Jain, and S. Prabhakar, Handbook of fingerprint recognition, 2nd ed. Springer Science & Business Media, 2009.

R. D. Labati, A. Genovese, E. Muñoz, V. Piuri, and F. Scotti, “A novel pore extraction method for heterogeneous fingerprint images using convolutional neural networks,” Pattern Recognition Letters, vol. 113, pp. 58–66, 2018.

H.-U. Jang, D. Kim, and S.-M. Mun, “Deeppore: fingerprint pore extraction using deep convolutional neural networks,” IEEE Signal Processing Letters, vol. 24, no. 12, pp. 1808–1812, 2017.

H. Wang, X. Yang, L. Ma, and R. Liang, “Fingerprint pore extraction using u-net based fully convolutional network,” in Biometric Recognition (CCBR). Springer International Publishing, 2017, pp. 279–287.

M. Ali, C. Wang, and M. O. Ahmad, “An efficient convolutional neural network for fingerprint pore detection,” IEEE Transactions on Biometrics, Behavior, and Identity Science, vol. 3, no. 3, pp. 332–346, 2021.

J. Cui, M.-S. Ra, and W.-Y. Kim, “Fingerprint pore matching method using polar histogram,” in Proceedings of the 2014 IEEE International Symposium on Consumer Electronics (ISCE). Hsinchu, Taiwan: IEEE, 2014, pp. 1–2.

A. B. V. Wyzykowski, M. P. Segundo, and R. d. P. Lemes, “Level three synthetic fingerprint generation,” arXiv preprint arXiv:2002.03809, 2020. [Online]. Available: [link]

O. Ronneberger, P. Fischer, and T. Brox, “U-net: Convolutional networks for biomedical image segmentation,” 2015. [Online]. Available: [link]