UMP Institutional Repository

Integration of median filter and oriented field estimation for fingerprint identification system

Nor'Aqilah, Misman (2015) Integration of median filter and oriented field estimation for fingerprint identification system. Masters thesis, Universiti Teknologi Malaysia.

[img]
Preview
PDF
NOR'AQILAH BINTI MISMAN.PDF

Download (805kB) | Preview

Abstract

The Fingerprint Identification System (FIS) has been used and applied into various aspects. The system used identification based on fingerprint to give an authorization and identification to every person that wants to access the system. However, there are some research issues that affect the system accuracy such as noise element and low-quality fingerprint image. To solve this problem, this project will proposed two selection methods; which are Median filter to reduce noise element and Orientation Field Extimation method to enhance the low quality image. This proposed methods is implement in order to get an accurate result and high performance system. In order to verify the system identification, two experiments has been done which are functional test and accuracy test. This test will used 16 images from FVC2004DB1 set. From this test, there will be three results that being focus on which are the computational time, high peak value, False Rejection Rate (FRR), False Acceptance Rate (FAR) and Matching Rate. These values are used in order to verify high performance in the system, by comparing the proposed system with other existing system. By doing this experiment, it shown that by using the proposed methods it has lower value in average time and FRR value; which is good in order to get a high performance working system. However, for FAR value the other existing work has more accurate result in identifying fingerprint image compared to proposed work. Based from the experimental test, it shown that by using the proposed methods it is effective in order to identify low-quality and noises image with an accurate matching result and high performance system.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Neural networks (Computer science); Fingerprints; Identification; Data processing
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Depositing User: Ms. 'Arifah Nadiah Che Zainol Ariff
Date Deposited: 17 Mar 2016 02:13
Last Modified: 17 Mar 2016 02:13
URI: http://umpir.ump.edu.my/id/eprint/12075
Download Statistic: View Download Statistics

Actions (login required)

View Item View Item