Image Quality Assessment for Fake Biometric Detection

Abstract

This paper presents Image quality assessment for fake biometric system. The key point of the process is to find a set of discriminant features which permits to build an appropriate classifier which gives the probability of the image “realism” given the extracted set of features. In the present work, we propose a novel parameterization using 25 general full referenced and non-referenced image quality measures. In order to keep its generality and simplicity, the system needs only one input: the biometric sample to be classified as real or fake. The work was carried using Iris (ATVS-Flr DB) and Fingerprint(Livedet09) datasets. The simulation results indicate a significant accuracy of 95% with Iris biometry and 92.5% from fingerprint.