Enhanced block filter thinning algorithm

Other Title(s)

تحسين خوارزمية ال (BFT)‎

Dissertant

al-Tarawinah, Aishah Zal Ahmad

Thesis advisor

al-Abbadi, Muhammad

University

Mutah University

Faculty

Information Technology College

Department

Computer Science Department

University Country

Jordan

Degree

Master

Degree Date

2016

English Abstract

Biometrics is an evolving technology in the society of today.

Fingerprint recognition technology is among the extensively utilized biometrics systems globally.

Thinning is a concept that is quite fundamental in the prior phases of automatically recognizing fingerprints.

Minutiae extraction performance is dependent substantially on the skeleton quality used in the identification process.

An exceptionally efficient algorithm for fingerprint thinning can enhance minutiae extraction and alleviate image noises therefore facilitating in enhancing system performance.

This dissertation involves research in EBFT (Enhance Block Filter Thinning) fingerprint matching scores.

A comparison on performance levels of three distinct thinning algorithms is made.

These are: CLT (Central line thinning), EBFT and BFT (Block Filter Thinning).

The system essentially examines all the existing possible dataset alignments and produces matching scores on the most distinct alignment.

The comparison of Algorithms is based on false match and the false non matching scores as well as the time that has elapsed in relation to each of the existing algorithms.

In the final stage a succession of matching trials is used to compare performance of the three previously highlighted algorithms i.e.

CLT, EBFT and BFT.

The results obtained indicate that EBFT thinning formula yields enhanced false match and false non matching rates compared to the rates obtained from BFT algorithm thinning.

BFT algorithms take the least time to process while CLT algorithms yield the most accurate false matching and false non matching rates although it takes a substantial amount of time.

CLT is a better method compared to the methods that this literature will examine.

The experiments performed in this research utilize synthetic fingerprint generator or SFingGe software.

This method is devoid of noise produced by the sensor.

In total and for the purpose of the database creation of 60 base images took place.

Main Subjects

Information Technology and Computer Science

No. of Pages

122

Table of Contents

• APA
• MLA
• AMA

Language

English

Data Type

Arab Theses

Record ID

BIM-749298