A New Approach for Fingerprint Verification Based on Wide Baseline Matching Using Local Interest Points and Descriptors
In this article is proposed a new approach to automatic fingerprint verification that is not based on the standard ridge-minutiae-based framework, but in a general-purpose wide baseline matching methodology. Instead of detecting and matching the standard
- PDF / 930,948 Bytes
- 14 Pages / 430 x 660 pts Page_size
- 56 Downloads / 199 Views
1 Introduction Fingerprint verification is one of the most employed biometric technologies. A fingerprint is the pattern of ridges and furrows on the surface of a fingertip. It is formed by the accumulation of dead, cornified cells [5]. The fingerprint pattern is unique and determined by the local ridge characteristics and the presence of ridge discontinuities, called minutiae. The two most prominent minutiae are ridge termination and ridge bifurcation. Minutiae in fingerprints are generally stable and robust to fingerprint impression conditions. Singular points, called loop and delta, are a sort of control points around which the ridge-lines are “wrapped” [12]. Many approaches to automatic fingerprint verification have been proposed in the literature and the research on this topic is still very active. In most of the cases the automatic verification process is based on the same procedure employed by human experts: (i) Detection of structural features (ridges, minutiae, and/or singular points), and in some cases derived features as the orientation field, which allow characterizing the fingerprints, and (ii) Comparison between the features in the input and reference fingerprints. This comparison is usually implemented using minutiae-based matching, ridge pattern D. Mery and L. Rueda (Eds.): PSIVT 2007, LNCS 4872, pp. 586–599, 2007. © Springer-Verlag Berlin Heidelberg 2007
A New Approach for Fingerprint Verification Based on Wide Baseline Matching
587
comparison and/or correlation between the fingerprints. The mentioned comparison methodologies can be described as [12]: Minutiae-based matching: It consists of finding the alignment between the input and the reference minutiae sets that results in the maximum number of minutiae pairings; Ridge feature-based matching: The approaches belonging to this family compare fingerprints in term of features extracted from the ridge pattern (e.g. local orientation and frequency, ridge shape, texture information); and Correlation-based matching: Two fingerprint images are superimposed and the correlation (at the intensity level) between corresponding pixels is computed for different alignments (e.g., various displacements and rotations). In state of the art fingerprint verification systems several structural features and comparison methodologies are jointly employed. For instance, in the 2004 Fingerprint Verification Competition (FVC2004) the 29 participants (from 43) that provided algorithm’s information employed the following methodologies [2]: • Features: minutiae (27), orientation field (19), singular points (12), ridges (10), local ridge frequency (8), ridge counts (6), raw or enhanced image parts (4), and texture measures (3). • Comparison methodology: minutiae global (20), minutiae local (15), correlation (7), ridge pattern geometry (5), and ridge pattern texture (2). In this general context the main objective of this article is to propose a new approach to automatic fingerprint verification that it is not based on the standard ridge-minutiae-based framework, but in a gene
Data Loading...
