Fingerprint Presentation Attack Detection Utilizing Spatio-Temporal Features

Abstract

This paper presents a novel mechanism for fingerprint dynamic presentation attack detection. We utilize five spatio-temporal feature extractors to efficiently eliminate and mitigate different presentation attack species. The feature extractors are selected such that the fingerprint ridge/valley pattern is consolidated with the temporal variations within the pattern in fingerprint videos. An SVM classification scheme, with a second degree polynomial kernel, is used in our presentation attack detection subsystem to classify bona fide and attack presentations. The experiment protocol and evaluation are conducted following the ISO/IEC 30107-3:2017 standard. Our proposed approach demonstrates efficient capability of detecting presentation attacks with significantly low BPCER where BPCER is 1.11% for an optical sensor and 3.89% for a thermal sensor at 5% APCER for both.

A comprehensive survey of fingerprint presentation attack detection

“…Husseis et al [59] used eight global measures that included intensity, contrast and randomness. These features were extracted from fingerprint videos.…”

“…Husseis et al [60] , in an extension of their previous work [59] , utilized videos of fingerprints to extract five spatiotemporal features that allowed them to detect PAIs. An SVM with a second-degree polynomial kernel was used for classification.…”

“…[59] 2020Own dataset Extraction of eight global measures that included intensity, contrast, and randomness BPCER of 18.1% at 5% APCER for thermal sensor BPCER of 19.5% at 5% APCER for optical sensor Husseis et al [60] 2021 Own dataset Extraction of five spatiotemporal features BPCER of 3.89% at 5% APCER for thermal sensor BPCER of 1.11% at 5% APCER for optical sensor FLR: False live rejection; FSA: false spoof acceptance; EER: equal error rate; BPCER: bona fide presentation classification error rate.…”

On the Effectiveness of Impedance-Based Fingerprint Presentation Attack Detection

“…For this purpose, the SVM [ 52 ] is used as a classifier, since it constantly achieved remarkable performances across different fingerprint PAD studies [ 34 , 35 , 53 , 54 , 55 , 56 , 57 ]. The SVM is designed to work on high-dimensional input data and derive binary decisions by defining a hyperplane that separates both classes.…”

The Impact of Pressure on the Fingerprint Impression: Presentation Attack Detection Scheme

“…Moreover, in the context of dynamic PAD mechanisms, we analyzed in our previous work [22] the variation of first order statistics in fingerprint videos of different sensing technologies where the mechanism achieved 18.1% BPCER for the thermal sensing and 19.5% BPCER for the optical sensing at 5% APCER for both. We extended the experiment by utilizing spatio-temporal feature extractors that consolidate the spatial fingerprint features with the temporal variations, which improved the accuracy to 3.89% BPCER for the thermal sensing and 1.11% BPCER for the optical sensing at 5% APCER for both [23].…”