Abstract:
Fingerprint matching is an important and challenging research area of Digital Image
Processing. Now a day’s, every country of the world is much more concerned about their
safety and security concerns than ever. That’s where a fingerprint verification system
helps in forbidding unauthorized access to different facilities, and in situations of any
breach it also helps in tracing the intruders.
Minutiae based approach is one of the most famous technique for fingerprints matching.
Minutiae are actually the features that are attained by the ridge discontinuities. Ridge
endings and bifurcation are most commonly used minutiae types and researchers have
used them in many flavors by using their attributes like Minutia type, minutia
coordinates, Distance between minutiae, Ridges count between minutiae, Direction and
relative angles etc. The performance of the minutiae based classification depends on the
strength of these features. Due to the noise or corruption in the image integrity of these
features reduces. Therefore it is required that the features extracted from these minutiae
should be robust enough in such a way that it can minimize the effects of problem caused
by some kind of noise and should have features from multiple domains.
The purpose of this research revolves around the fact described in the previous passage.
Efforts are made to extract a very rich feature sets that covers not only features from
spatial domain but also some features are extracted from frequency domain after
applying different wavelets. And remarkably for classification using features that are
actually attained by fusing both types of features really helped and certainly improved
system performance. The system was tested on standard fingerprint database and very
good results are obtained.
