Multimodal Liveness Detection System for Presentation Attacks on Fingerprint and Iris Biometrics

Abstract

Biometric-based person identification is gaining its importance and now-a-days it is newlinerequired to process a large amount of biometric data. In general, biometric authentication newlinecan vulnerable to spoofing attacks. Spoofing referred to presentation attack newlineto mislead the biometric sensor. An anti-spoofing method is able to automatically newlinedifferentiate between real biometric traits presented to the sensor and synthetically newlineproduced artifacts. Further, the problem is compounded when we have to deal with newlinetwo or more biometric traits. The objective of this thesis is to address these issues newlineand investigate the liveness detection for unimodal and multimodal biometric traits. newlineIn this thesis, an efficient software-based liveness detection method is proposed newlinethat can classify the fake and real biometric traits. A novel combination of local newlinemicro-texture features and macro-texture features derived from GLCM and NGTDM newlinerespectively is used to generate an effective feature vector. Support vector machine newlineclassifier is used for discriminating between live and fake fingerprints. newline