Abstract: Diabetic Retinopathy (DR) is an eye disease due to diabetes which causes visual loss. So it is reasonable to provide treatment for DR at first stages of disease. Hemorrhages are the first symptoms that indicate diabetic retinopathy. Therefore, their detection is very important. In this survey paper image contrast is enhanced by pre-processing and then blood vessels are detected as the boundaries of hemorrhages are not preserved when they are in contact with blood vessels.then classification done on the basis texture feature such as area and size and standard deviation etc. finally classify the images into normal, moderate and Severe DR.

Keywords: Diabetic retinopathy, haemorrhages, Blood vessel, Diabetes.