Diabetes is increasing at a faster rate. Due to diabetes various diseases are caused. Diabetic retinopathy is one of the major complications caused due to diabetes that affects the retina of a eye. Using fundus image analysis Diabetic Retinopathy can be detected for which feature extraction algorithms are implemented on the fundus images. Earlier the process of detecting diabetic retinopathy based on abnormal features like exudates and hemorrhages was time consuming. Latest technological advances help in automatic detection of diabetic retinopathy. Objective of the project is feature detection of fundus images. In the study, a blob detection methodology is proposed for detecting the features such as macula, exudates and hemorrhages.

Keywords : Diabetic retinopathy, fundus images, feature detection, macula, exudates, hemorrhages.