Title

Fingerprint Identification Using Delaunay Triangulation

Keywords

Delaunay triangulation; Fingerprint recognition; Indexing

Abstract

Presents an indexing-based approach to fingerprint identification. Central to the proposed approach is the idea of associating a unique topological structure with the fingerprint minutiae using Delaunay triangulation. This allows for choosing more "meaningful" minutiae groups (i.e., triangles) during indexing, preserves index selectivity, reduces memory requirements without sacrificing recognition accuracy, and improves recognition time. Specifically, assuming N minutiae per fingerprint on average, the proposed approach considers only O(N) minutiae triangles during indexing or recognition. This compares favorably to O(N3), the number of triangles usually considered by other approaches, leading to significant memory savings and improved recognition time. Besides their small number, the minutiae triangles we used for indexing have good discrimination power since, among all possible minutiae triangles, they are the only ones satisfying the properties of the Delaunay triangulation. As a result, index selectivity is preserved and indexing can be implemented in a low-dimensional space. Some key characteristics of the Delaunay triangulation are: (i) it is unique (assuming no degeneracies), (ii) can be computed efficiently in O(NlogN) time, and (iii) noise or distortions affect it only locally. The proposed approach has been tested on a database of 300 fingerprints (10 fingerprints from 30 persons), demonstrating good performance.

Publication Date

1-1-1999

Publication Title

Proceedings - 1999 International Conference on Information Intelligence and Systems, ICIIS 1999

Number of Pages

452-459

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

DOI Link

https://doi.org/10.1109/ICIIS.1999.810315

Socpus ID

84902322254 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/84902322254

This document is currently not available here.

Share

COinS