Design and Development of an Automated System for Early Detection of Diabetic Retinopathy using Image Processing

Abstract

Diabetic retinopathy is a disorder that accounting as a basic origin of uncorrectable vision loss for diabetic patients. The difference in blood vessels of retina leads to a most general disease called diabetic retinopathy. However, earlier identification and remedy through regular screening, blindness can be avoided. Exudates are prime sign of diabetic retinopathy that can feasibly be measured automatically. The focus is to evolve an automated algorithm to analyse the fundus images, perform image processing to extract hard exudates which are fundamental indications of diabetic retinopathy. Hard exudates are small, yellow in color waxy patches of white color. This project implements a method or algorithm which is fast and efficient to detect the exudates. Identify the feature of exudates from the image techniques using CBIR database and to prescribe proper medication.