Research Article
Adaptive Thresholding Technique for Retinal Vessel
Segmentation Based on GLCM-Energy Information
Temitope Mapayi,
1
Serestina Viriri,
1
and Jules-Raymond Tapamo
2
1
School of Mathematics, Statistics & Computer Science, University of KwaZulu-Natal, Durban 4000, South Africa
2
School of Engineering, University of KwaZulu-Natal, Durban 4041, South Africa
Correspondence should be addressed to Jules-Raymond Tapamo; tapamoj@ukzn.ac.za
Received 9 August 2014; Revised 24 November 2014; Accepted 24 November 2014
Academic Editor: Lev Klebanov
Copyright © 2015 Temitope Mapayi et al. his is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
Although retinal vessel segmentation has been extensively researched, a robust and time eicient segmentation method is
highly needed. his paper presents a local adaptive thresholding technique based on gray level cooccurrence matrix- (GLCM-
) energy information for retinal vessel segmentation. Diferent thresholds were computed using GLCM-energy information. An
experimental evaluation on DRIVE database using the grayscale intensity and Green Channel of the retinal image demonstrates
the high performance of the proposed local adaptive thresholding technique. he maximum average accuracy rates of 0.9511 and
0.9510 with maximum average sensitivity rates of 0.7650 and 0.7641 were achieved on DRIVE and STARE databases, respectively.
When compared to the widely previously used techniques on the databases, the proposed adaptive thresholding technique is time
eicient with a higher average sensitivity and average accuracy rates in the same range of very good speciicity.
1. Introduction
Retinal fundus imaging in ophthalmology is of great use
in medical diagnosis and progression monitoring of several
diseases like hypertension, diabetes, stroke, and cardiovas-
cular disease [1]. Automatic vessel segmentation has a great
potential to assist in the reduction of the time required by
physicians or skilled technicians for manual labeling of retinal
vessels [2].
Several retinal vessel segmentation techniques have been
proposed and evaluated in literatures. Chaudhuri et al.
[3] implemented a two-dimensional matched ilter using a
Gaussian shaped curve. However, the technique proposed in
[3] achieved very low average accuracy due to low detection
of retinal vessels. Hoover [4] segmented retinal vessels by
applying a threshold probing technique combining local ves-
sel attributes with region-based attributes on matched ilter
response (MFR) image. When compared to [3] where a basic
thresholding of an MFR was used, the method proposed by [1]
reduced the false positive rate by as much as 15 times. Fraz et
al. [5] implemented vessel segmentation technique utilizing
extracted center-lines of retinal vessels through irst-order
derivative of Gaussian ilter. he authors used morphological
operator with directional structuring elements to enhance the
structure of blood vessels. hey further generated the shape
and orientation map of the blood vessels using the bit planes
of a gray-scale image. Chakraborti et al. [6] implemented
an unsupervised segmentation technique that combines ves-
selness ilter and matched ilter using orientation histogram
for the segmentation of retinal vessels. Martinez-Perez et al.
[7] used a combination of scale space analysis and region
growing to segment the vasculature. he technique proposed
in [7] was however unable to segment the thin vessels. Zana
and Klein [8] used a general vessel segmentation method
based on mathematical morphology. However, the technique
proposed in [8] was unable to segment the thinner vessels.
Wang et al. [9] proposed multiwavelet kernels and mul-
tiscale hierarchical decomposition. Vessels were enhanced
using matched iltering with multiwavelet kernels. Szpak and
Tapamo [10] used gradient based approach and level set tech-
nique. he proposed technique in [10] was however unable
to detect the thinner vessels. Vlachos and Dermatas [11]
proposed a multiscale retinal vessel segmentation method.
he algorithm is based on multiscale line-tracking procedure
Hindawi Publishing Corporation
Computational and Mathematical Methods in Medicine
Volume 2015, Article ID 597475, 11 pages
http://dx.doi.org/10.1155/2015/597475