An Enhanced Iris Segmentation Algorithm Using Circle Hough Transform

Nkole, Ifeanyi Ugbaga and Sulong, Ghazali Bin and Saparudin, Saparudin (2012) An Enhanced Iris Segmentation Algorithm Using Circle Hough Transform. In: 2nd Basic Science International Conference (BaSIC) 2012, 24-25 Februari 2012, Malang.

[thumbnail of Confer@Indo.pdf]

Download (871kB) | Preview
[thumbnail of Confer@Indo.docx] Text
Restricted to Registered users only

Download (326kB) | Request a copy


Iris segmentation is the most contesting issue in the iris recognition system, since the result of this stage will mar or break the iris recognition system effectiveness. Therefore, a very careful attention has to be paid in the segmentation process if only an accurate result is expected; this depends on the accuracy of the detected pupil center. In this paper we proposed a new method to localize the center of the pupil which is concentric with the iris image by employing 8-neighbourhood operators. This parameter is then fed to a Circle Hough Transform to enhanced iris segmentation processing speed and accuracy. The occlusions due to eyelids and eyelashes noise are detected by applying canny edge operator. The experiment is conducted using 320 iris images from CASIA standard dataset, and the result shows that the proposed method had a high accuracy rate.

Item Type: Conference or Workshop Item (Paper)
Additional Information: Saparudin-Ph.D Candidate, Lecturer of Computer Science Sriwijaya University
Uncontrolled Keywords: Iris segmentation, Iris recognition, 8-neighbourhood operator, Circle Hough transform, and Canny edge detection.
Subjects: Q Science > Q Science (General) > Q1-295 General
Q Science > QA Mathematics > QA75-76.95 Calculating machines > QA75 Electronic computers. Computer science
Q Science > QA Mathematics > QA75-76.95 Calculating machines > QA76 Computer software
Divisions: 09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User: backup admin
Date Deposited: 09 Jan 2020 07:10
Last Modified: 09 Jan 2020 07:10

Actions (login required)

View Item View Item