Publications

Abstract : Advancement in technology has increased the need to provide adequate security. Hence came the first security system using passwords, these systems have evolved and started using biometrics traits instead of passwords. But people have found ways to cheat such systems through different means. There have been many methods to prevent this security vulnerability, but these methods are very complex or not easy to implement. The most commonly used biometric trait for verification and also the most spoofed biometric trait is fingerprint. The need is to find an easy way to detect a fake fingerprint. The proposed system uses fingerprint and the photo of the fingernail when light is passed through is taken. The image of the fingernail is processed and features are extracted using SIFT and compared with feature vector stored in the database. The correlation is then calculated. Based on the correlation the system decides whether the fingerprint is fake or not. The system was found to react positively to intensity changes in the fingernail image and difference in the pattern of the fingernail. This system provides a simple and robust method for detection of fingerprint spoofing.