A Novel Approach to Fingerprint Identification

Abstract

In biometrics fingerprint is the widely and oldest technology used as identification. In this thesis a database is has been created from the extracted minutiae points. Database is designed using B Tree indexing techniques. By using this technique there is no need of comparing fingerprint image with other fingerprint. The B Tree algorithm further works on the computational algorithm based on minutiae points distance calculations. Moreover, it is known that everyone has unique fingerprint, but no mathematical proof has been presented which can present as it’s hard to match some fingerprint patterns with whole world’s patterns. In this paper a mathematical proof on the basis of adjacent minutiae points distance calculation has been presented. This calculation shows that every fingerprint is unique. Further by using extracted minutiae a database has been designed. As no actual image comparison is presence hence it increase the searching speed and decrease the database size.