Minutiae-Based Fingerprint Identification

Abstract

Minutiae-based method is the most popular and widely used technique in personal identification, criminal investigation and being the basis of the fingerprint comparison made by fingerprint examiners. This technique, however, still suffers from the problem of poor quality image. Marking all the minutiae accurately as well as rejecting the false minutiae is another issue which is still under research. My work has combined many methods to build minutia extractor and minutia matcher. In this thesis, I propose a method in which we find out the maximum angle of the ridges by measuring the exact value of number of rows and columns of ridges to improve the quality of orientation field based on well-known excel statistical function – the

LINEST. Moreover, this improved method to orient the fingerprint image combined with the Poincare index method will help us to detect the core and delta points which links to minutiae such ridge ending and ridge bifurcation to create the concise feature vectors used for matching.