A Novel Method of Fingerprint Recognition Based on Neural Network

Abstract: Fingerprints are the great way to identify individuals. Biometric identification through fingerprint is one of the oldest methods. To get a good fingerprint is not always easy because of injury on an individual�s finger and noisy image. So pre-processed of the fingerprint image is necessary before matching. The main aim of this work is to get better and enhanced fingerprint image. We have analysed the components to obtaining high-performance feature point�s detection algorithm, such as quality of image, separation, enhancement of image and feature detection. It is necessary to accurate Separation of noisy background from fingerprint ridges. The simulations are completed in the MATLAB2014 environment to evaluate the performance of the applied algorithms.

Keywords: Biometric, Fingerprints, Neural Network, Feedforward network, MATLAB2014.