Bad quality fingerprint reconstruction method using Gabor filtering and connected component analysis

Abstract

Reconstruction of bad quality fingerprint is a common and critical step in fingerprint recognition systems as it provides the base for the performance of the system. In India fingerprints are widely used in order to prove a person’s unique identity. Despite the widespread use of fingerprints, there is little statistical theory on the uniqueness of fingerprint minutiae. In this work, first work has been done enhancement in which input fingerprint is divided into two phases called spiral phase and continuous phase. First the image is enhanced using spiral and continuous phase regions with the help of gabor filter and connected component analysis in which only quality pixels are kept and other are excluded and then matching has been carried out to see the comparison with and without reconstruction. Results show improved performance of the algorithm after reconstruction of bad quality regions in the original input fingerprint images