A Pore-Based Method for Fingerprint Liveness Detection

Fingerprint scanners usually suffer from spoofing attack of fake fingerprints. To address this issue, a pore based liveness detection method has been proposed in this paper. The method first extracts the sweat pores from the grayscale fingerprint image. This is achieved by processing the fingerprint image using the Gaussian filters, whose parameters are adaptively adjusted according to the local ridge period of fingerprints. Based on the extracted pores density feature, we then find an optimal threshold of the pores density to distinguish between the live and fake fingerprints. In the experiments, we evaluate the proposed method on ATVS-FFp database. The results show that our method is robust, achieving an average classification error rate (ACE) at 7.62%. Furthermore, the proposed method is simple and easy to be implemented, as it requires only one fingerprint image with 500 dpi for liveness detection.