A hybrid partial fingerprint matching algorithm for estimation of Equal error rate

Abstract

In Biometric fingerprints are most promising technology for identification and verification in various applications. The general fingerprint matching algorithms perform poorly in case of partial prints and may discard these prints during matching. The major problems with partial prints is the lack of level 1 and level 2 features which makes them distinguishable. Therefore, we can utilize level 3 features such as pores corresponding LBP ectraction in combination with level 2 features based radon transform. Pores are one of the discriminative level 3 feature and with the advancement in the technology they can be successively employed using Automatic Fingerprint (AFIS) identification systems. In this paper, Our main aim is to provide a method to increase accuracy of partial fingerprint matching in order to estimate Equal error rate(ERR) based on False Acceptance Rate and False Rejection Rate. In proposed method, score level fusion of minutiae based radon transform and pores based LBP matching is performed. This method can surpasses the results of other matching approach which uses single matching sceme. The performance can be evaluated by calculating ERR based on False Acceptance Rate (FAR) and False Rejection Rate (FRR).