Fingerprint Spoofing Detection Using Machine Learning

Abstract

Many studies had used different techniques to produce liveness fingerprint detection schemes. Our paper aims to review the different studies proposed in liveness fingerprint detection systems that able to classify real fingerprint images and fake ones using machine learning techniques, and to analyze different schemes. A comparison between datasets used in the literature was made based on specific metrics. The results indicate that the most extracted features are BSIF, LPQ. Support vector machine algorithm SVM were widely used as classifier.