Page 26 - 2018

Ph.D.
(Computer Applications)
ISSUES AND PERFORMANCE IMPROVEMENT FOR
BIOMETRIC SYSTEM

Ph.D. Scholar : Patel Meghnaben Babubhai
Research Supervisor : Dr. A. R. Patel, Co-Guide : Dr. S. M. Parikh

Regi. No.: 11146041005
Abstract :
In the era of information security biometric is the most secure and appropriate method
for encouraging the person recognition and identification. It authenticate a person based
on physiological (material or behavioral) characteristics of the person.
Fingerprint authentication is the most famous and widely adopted biometric over a
century because it remain unique and consistent over time. It is popular because used in
many sectors like government, national and international security, law enforcement
agencies and so on for making safe and secure world.

Many different types of algorithms and sensors available in the market and to develop the
fingerprint standard is become an essential element in fingerprint recognition in both
national and international levels. The image obtain by one device is interpreted by another
device means interoperability play a vital role in product implementation.
Even though the widely use of fingerprints, still some less statistical work done on the
uniqueness of fingerprint minutiae. Still the how many minutiae pair should be match is
unresolved for fingerprint matching. The proposed research work focus on following
major challenges which are exist in fingerprint recognition.

First major challenge in fingerprint recognition system is poor quality images. The
performance of the fingerprint recognition is influenced by the quality of an image. To,
overcome this problem at initial level use the normalization method which remove the
noise which is created by the pressure given on scanner. As well as at image
enhancement phase proposed research used the O’Gorman filter and also enhanced it for
improving the quality of an input image. The performance is also compared with most
famous image enhancement method called as Gabor filter. The result prove that
enhanced O’Gorman filter connect broken ridges, remove noises between the ridges and
improve the ridge and valley contrast.
Second major challenge in fingerprint recognition system is displacement of fingerprint
which give false result at the time of fingerprint matching. To overcome this problem
used orientation estimation which represent the direction of ridge flow. It is mainly used
for ridge enhancement as well as for singular detection. The proposed research work

07