UTHM Institutional Repository

An improved low contrast image in normalization process for iris recognition system

Ghali, Abdulrahman Aminu and Jamel, Sapiee and Mohamad, Kamaruddin Malik and Ahmad Khalid, Shamsul Kamal and Pindar, Zahraddeen Abubakar and Mat Deris, Mustafa (2018) An improved low contrast image in normalization process for iris recognition system. In: Advances in Intelligent Systems and Computing. Recent Advances on Soft Computing and Data Mining, 700 . Springer, 495-505 . ISBN 9783319725505

Full text not available from this repository.

Abstract

Iris recognition system is one of the most predominant methods used for personal identification in the modern days. Low quality iris image such as low contrast and poor illumination presents a setback for iris recognition as the acceptance or rejection rates of verified user depend solely on the image quality. This paper presents a new method for improving histogram equalization technique to obtained high contrast in normalization process thereby reducing False Rejection Rate (FRR) and False Acceptance Rate (FAR). The proposed technique is developed using C++ and tested using four datasets CASIA, UBIRIS, MMU and ICE 2005. The experimental results show that the proposed technique has an accuracy of 95%, as compared to the existing techniques: CLAHE, AHE, MAHE and HE which have an accuracy of a 93.0, 85.7, 92.8 and 90.71% respectively. Hence it can be concluded that the proposed technique is a better enhancement technique compared to the existing techniques for image enhancement.

Item Type: Book Section
Uncontrolled Keywords: Iris recognition; histogram equalization; image enhancement; normalization
Subjects: Q Science > QA Mathematics > QA76 Computer software
Divisions: Faculty of Computer Science and Information Technology > Department of Information Security
Depositing User: Mr. Mohammad Shaifulrip Ithnin
Date Deposited: 31 Jul 2019 00:59
Last Modified: 31 Jul 2019 00:59
URI: http://eprints.uthm.edu.my/id/eprint/11368
Statistic Details: View Download Statistic

Actions (login required)

View Item View Item