Fingerprint Recognition and feature extraction using transform domain techniques

Abstract

Fingerprint Recognition is one of the oldest and popular methods for person identification. There are two major approaches of fingerprint recognition namely image based and minutiae based. This paper presents an image based fingerprint recognition method. In this work first core point is detected and pre processing of image is done then transformation technique like Fast Fourier Transform (FFT), Discrete Cosine Transform (DCT), Discrete Wavelet Transform (DWT) and Gabor Filter are used for feature extraction. These features consist of mean energy, standard deviation and Shannon entropy. The good recognition accuracy is achieved on the FVC 2002 database. Comparison of all the four transform is presented here and it is observed that DCT and DFT gives better result as compared to DWT and Gabor filter.