Adaptive thresholding pattern for fingerprint forgery detection
Zahra Farzadpour, Masoumeh Azghani
TL;DR
The paper addresses fingerprint forgery detection under distortions in a software-based setting. It introduces a detector that fuses anisotropic diffusion ($AD$) with a three-level Haar wavelet transform and an adaptive thresholding pattern ($ATP$) to create rich feature vectors, subsequently classified by an $SVM$. Thresholds per subband are computed from a reference image using $T^k(x,y)=\beta T^0(x,y) \exp(-\alpha(x,y)(k-1))$, with $T^0(x,y)=H(x,y)-L(x,y)$ and $\alpha(x,y)=H(x,y)/T^0(x,y)$. On ATVS-FFp DB, the method shows superior robustness to random pixel missing, block missing, and AWGN distortions, achieving near-0.97 accuracy under strong noise while outperforming several benchmarks under missing data scenarios.
Abstract
Fingerprint liveness detection systems have been affected by spoofing, which is a severe threat for fingerprint-based biometric systems. Therefore, it is crucial to develop some techniques to distinguish the fake fingerprints from the real ones. The software based techniques can detect the fingerprint forgery automatically. Also, the scheme shall be resistant against various distortions such as noise contamination, pixel missing and block missing, so that the forgers cannot deceive the detector by adding some distortions to the faked fingerprint. In this paper, we propose a fingerprint forgery detection algorithm based on a suggested adaptive thresholding pattern. The anisotropic diffusion of the input image is passed through three levels of the wavelet transform. The coefficients of different layers are adaptively thresholded and concatenated to produce the feature vector which is classified using the SVM classifier. Another contribution of the paper is to investigate the effect of various distortions such as pixel missing, block missing, and noise contamination. Our suggested approach includes a novel method that exhibits improved resistance against a range of distortions caused by environmental phenomena or manipulations by malicious users. In quantitative comparisons, our proposed method outperforms its counterparts by approximately 8% and 5% in accuracy for missing pixel scenarios of 90% and block missing scenarios of size 70x70 , respectively. This highlights the novelty approach in addressing such challenges.