Comparative Study in Classifying Fake Fingerprint Images

Abstract

Biometrics is the measurement for body calculation and statistical analysis. It is commonly used as form of security technology. Fingerprints recognition is used as one of the biometrics due to its strong conditions as each individual has a unique fingerprint pattern. Unfortunately, it is also exposed to every type of attacks. Fake fingerprint in biometric devices. The biometrics devices needed to have a software algorithm to classify between real or fake fingerprint to combat breach of security that used biometrics. Due to this, liveness detection comes to light to differentiate live and fake fingerprints. The proposed method in this study is using Support Vector Machine and Naïve Bayes as a classifier to compare which classifier can produce the best accuracy on detecting the fake fingerprint. The classifiers will undergo few cross-validations that will be presented in this study to produce which classifiers have better accuracy on detecting fake fingerprints.