Retinal vessel segmentation using a modified morphology process and global thresholding

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Citations (Scopus)

Abstract

The retinal blood vessels are one of the objects in the retina. The characteristics of the retinal vessels have their unique shape and size. Diseases suffered by humans, including diabetic retinopathy, hypertension, and cardiovascular disorders, can be detected by examining these blood vessels. An automated system that can distinguish blood vessels from the other objects in the retina is here presented. It consists of preprocessing, segmentation, and accuracy calculations. Preprocessing consists of green channel, overlap masking, opening and top-hat morphology. Segmentation uses global thresholding. Experiments used the DRIVE image fundus dataset. Segmentation separates the blood vessel from the background. The performance of the segmentation results is measured in terms of accuracy, sensitivity and specificity. The average accuracy is 96.54%, the average specificity is 98.81%, and the average sensitivity is 72.27%.

Original languageEnglish
Title of host publication8th Annual Basic Science International Conference
Subtitle of host publicationCoverage of Basic Sciences toward the World's Sustainability Challenges
EditorsCorina Karim, Rodliyati Azrianingsih, Mauludi Ariesto Pamungkas, Yoga Dwi Jatmiko, Anna Safitri
PublisherAmerican Institute of Physics Inc.
ISBN (Electronic)9780735417397
DOIs
Publication statusPublished - 17 Oct 2018
Event8th Annual Basic Science International Conference: Coverage of Basic Sciences toward the World's Sustainability Challanges, BaSIC 2018 - Malang, East Java, Indonesia
Duration: 6 Mar 20187 Mar 2018

Publication series

NameAIP Conference Proceedings
Volume2021
ISSN (Print)0094-243X
ISSN (Electronic)1551-7616

Conference

Conference8th Annual Basic Science International Conference: Coverage of Basic Sciences toward the World's Sustainability Challanges, BaSIC 2018
Country/TerritoryIndonesia
CityMalang, East Java
Period6/03/187/03/18

Keywords

  • Dilation
  • erosion
  • image fundus
  • retinal blood vessel

Fingerprint

Dive into the research topics of 'Retinal vessel segmentation using a modified morphology process and global thresholding'. Together they form a unique fingerprint.

Cite this