Investigation on feature based detection of glaucoma in fundus images

Abstract

Glaucoma is recognized as the second leading eye disorder that results in irreversible loss of vision. It is a retinal disorder that produces increased Intra Ocular Pressure (IOP) due to an imbalance in the production and drainage of the fluid within the eye. This increased IOP causes slow damage to the optic nerve fibre that carries the visual information from the retina to the brain. Early diagnosis and routine screening of eye help detection of the disorder and thereby prevents the irreversible loss of the vision. Computer-aided solutions help overcoming the challenges in the manual screening for a large population. Automated computed based screening tool helps reduction in the time and pressure of the ophthalmologist. The focus of this work is on the investigation of the potential in retinal image analysis for the detection of Glaucoma. newlineThe computer-based analysis of the parameter involves the use of image processing algorithms for pre-processing, localization and segmentation of the region of interest (ROI), extracting the features and computation of the parameters. The detection of Glaucoma done on the basis of the structural deformation demands the accuracy of segmenting the cup region and disc region. As a solution to this problem, non-morphological features are considered in this work for detecting the presence of Glaucoma. The non-morphological features extracted from spatial and transform domain include texture, intensity, histogram, and fractal. The initial step in the retinal image analysis involves the enhancing the contrast of the fundus image from the three database, Drishti-GS1, FAU and RIMONE. The retinal image is enhanced using Adaptive Intensity Transformation Contrast Enhancement, Adaptive Gamma Correction with Weighted Distribution, Sub-Histogram Equalization and Dynamic Stochastic Resonance. newline newline