Detection and analysis of retinal vasculature using optimization based multi level thresholding

Abstract

Analysis of blood vessels in digital retinal fundus images is an important task attempted in contemporary biomedical engineering research newlineIn this work normal and abnormal retinal images are pre processed with illumination compensation adaptive histogram equalization and fuzzy filtering Pre processed images are then treated with multi level thresholding methods namely Otsu and Tsallis Analysis using various statistical and Tamura features are carried out to determine the better method Further the chosen method is combined with optimization procedures such as particle swarm and bacterial foraging based techniques in order to improve the vessel content The obtained results are validated using similarity measures by comparing with the corresponding ground truth of each image Eight different features are derived from optimal multil evel thresholding output images to analyse the healthy and pathological images newlineResults demonstrate that attempted series of pre processing techniques enhances the edge information considerably and improves the efficacy of segmentation newline newline