Extraction of Different Features to Detect Diabetic Retinopathy from Retinal Fundus Image: A Review

Usually because of less secretion of insulin from the pancreas, the body is incapable of holding the process probably which directly identifies diabetes that is a chronic end organ disease. Diabetes can affect the whole circulatory system even including the retina. In case of Diabetic Retinopathy, retina is damaged because of fluid leaks from blood vessels into the retina. According to ophthalmologists, some basic features are there to recognise diabetic retinopathy such as blood vessel area, exudates, haemorrhages, microaneurysms and texture. In this paper some reviewed algorithms and implemented methods are included for extraction of these features from digital fundus image. Furthermore some classifications efficiency of different DR that use these features to classify individual fundus images are discussed. According to analysed fundus images, most reported systems are highly optimized for which generalization of individual result is difficult. This review shows that classification result is getting closure to classification capabilities of human ophthalmologists.