The 4th International Fuzzy Systems Symposium, İstanbul, Turkey, 5 - 06 November 2015, vol.1, no.1, pp.255-259
Retinal blood vessel segmentation is an important step for diagnosis of retinal diseases. An automatic vessel extraction technique can help specialist for this purpose. In this paper a new method is developed to extract vessel structures in retinal fundus images. Firstly, an enhancement procedure including Gabor Filter is performed to increase contrast between vessels and background after a preprocessing step. Afterward, top-hat transform is applied to gabor filter response
in order to make blood vessels more accurate. The output of the transformations is converted to binary image with a modified Fuzzy C-means clustering method. The images obtained from STARE and DRIVE databases which are available online are
used in order to test the developed system. A promising result is finally reached which are 95.63% and 95.80% of accuracy for STARE and DRIVE database respectively.