|
Title:
|
EXTRACTING PUPIL CONTOUR WITH HOUGH TRANSFORM OF CONNECTED COMPONENT BOUNDARIES |
|
Author(s):
|
Ivan A. Matveev, Nikolay N. Chinaev, Vladimir P. Novik |
|
ISBN:
|
978-989-8533-38-8 |
|
Editors:
|
Katherine Blashki and Yingcai Xiao |
|
Year:
|
2015 |
|
Edition:
|
Single |
|
Keywords:
|
Iris Segmentation, Hough Transform. |
|
Type:
|
Short Paper |
|
First Page:
|
275 |
|
Last Page:
|
280 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
A method for determining the pupil boundary in the image of eye is proposed. The method is based on examining connected components of binarized image. The components are first sorted according to probability to contain pupil, which is estimated from their inertia moments by empirically derived rules. Hough transform is applied to border of the component and outlines the center of most pronounced circular arc. If center hypothesis is strong enough the histogram of distances of border points to this center is built. If pronounced maximum exists on the histogram it means that a circle or an arc is present in this component, which is taken for pupil boundary. The method benefits from both stability of Hough approach and small number of pixels detected as boundaries of connected components. Tests were performed with several iris image databases from public domain. |
|
|
|
|
|
|
