Paradigm Shift in Fingerprint Recognition on Pressure Variation and Impact of Information System in Crime Reduction

Abstract

Fingerprint identification from a random offline fingerprint image has become a very actively studied field in biometric systems. This paper investigates the possibility to prepare a framework of applying the latent fingerprint obtained from a crime scene to be shared in a central data base. These fingerprints may be extremely useful to reduce the crimes from a country. In this approach each citizen is supposed to put his/her fingerprint in an organization during entry. This fingerprint will be matched to the central database for checking whether the fingerprint of that citizen matched with any finger print taken from a crime scene. The information system of an organization thus will be able to reveal the suspicious citizens causing to be investigated for further action. The quality of the fingerprint images greatly affects the performance of the minutiae extraction. The photo graph taken from different distances, uneven surface, different finger pressure, dust particles pose problems during recognition process. Improper finger pressure and uneven surface are the major cause to produce breaks in curves in a fingerprint. In order to improve the performance of the system, many researchers have been made efforts on the image enhancement algorithms. Most of the fingerprint recognition algorithms are based on minutia matching features. Therefore, minutiae extraction is one of the important steps in fingerprint verification algorithms. In this paper we present an algorithm to fill the broken curves on a fingerprint due to low finger pressure and uneven surface. Our proposed approach eliminates false minutiae that connect broken curves in fingerprint.