Extraction blood vessels from retinal fundus image based on fuzzy C-median clustering algorithm

Abstract

An automated method detection and extraction of blood vessels in retinal images would be described. The proposed algorithm is composed of three steps: matched filtering, fuzzy c-median (FCMED) clustering and label filtering. First, the matched filter technique is to enhance visualization of the blood vessels in retinal image. Secondly, the FCMED clustering is to keep the spatial structure of vascular tree segments. Finally, label filter technique is used to remove misclassified pixels. The algorithm has been tested on twenty sets in ocular fundus images, and experimental results are compared with clinically generated vessel segmentation and are evaluated in terms of sensitivity and specificity. This method performs well in analyzing anatomical structures in retinal image. © 2007 IEEE.