The Research Of Spoof Fingerprint Detection

Fingerprint authentication is one of the biometric authentication methods, and it is a powerful way to guarantee the safety of the financial and trade system, national security, e-commerce, etc. Spoof fingerprints attacks as a simple and practical way to crack a fingerprint recognition system was illegal used by certain criminals, which caused a huge threat to some social security system. For this emerging criminal means, the mainstream software-based method is texture analysis, but the original texture analysis doesn’t include coarseness analysis caused by the difference between the materials of spoof fingerprint and real fingerprint, and hardware-based method need for additional hardware, higher costs.A novel method is proposed based on the Curvelet transform and image texture feature with SVM-KNN classification, it also developed a fingerprint authentication system based on spoof fingerprint detection:Firstly, extracting Curvelet coefficient features with different scale and direction, analyzing the noise distribution of fingerprint images;Secondly, texture features are extracted from first order statistics, gray level co-occurrence matrix(GLCM) and Markov random field(MRF) in Curvelet reconstructed image;Lastly, fingerprint images are trained to obtain the classification criterion by SVM., and then SVM-KNN classification is used for spoof fingerprint detection;Moreover, taking the fake fingerprint detection methods into the fingerprint authentication system can prevents individual staff using a general vulnerability to login by spoof fingerprint detection, resulting in unpredictable losses and ensure the security of fingerprint authentication.Curvelet coefficient features and texture features are extracted to quantify discrimination, integration of texture and noise analysis, and this algorithm supports various types of fingerprint scanner. The experimental results in the databases of the Liveness Detection Competition2011(LivDet2011) have shown that the proposed method is effective and superior in spoof fingerprint detection.