Fingerprint Detection Using Minutiae Extraction

Abstract

In biometric application, identification of real or fake fingerprint is important for security purpose.It is mainly for the concern of fingerprint safety in authentication system.Ploy-Doh, silicon or other artifacts are used for making the fake fingerprints.The local feature descriptors are used for detecting the fingerprint but it is not satisfying the real world application. Based on the new trend convolutional neural network is used. It provides the better optimization process for both feature extraction and classifier training. Local binary pattern and minutiae extractions are used as a texture descriptor. These texture descriptors are used to identify the accuracy in the fingerprint image. Local binary pattern is used to convert the grey scale image into a binary image. It will check the accuracy of the fingerprint based on the 3x3 matrices pattern. Minutiae checks the ridge and bifurcation by following the process of binarization and thinning. Later the fusion algorithm is used to fuse both LBP and minutiae.