Segmentation of red lesions in diabetic retinopathy images

Diabetic retinopathy (DR) is a chronic disease which causes visual impairment in industrialized countries. Automatic detection of DR early markers can contribute to the diagnosis and screening of the disease. The aim is to automatically detect early signs of red lesions. So the need arises for fast and efficient system which detect whether the patient is suffering from diabetic retinopathy or not. Structure and location of blood vessels and optic disk play important role in accurate detection and classification of dark and bright lesions for early detection of DR. The developed method detect only the red lesions, by using Gabor filter, morphological Top-hat transformation, and entropy filter for the same then by combining the value of Gabor filter response and entropy filter response the red lesions can be obtained and from the subtraction of top-hat value from the obtained red lesions the spurious regions can be removed. Then its performance is evaluated, the same is done by the matched filter instead of Gabor filter for the blood vessel segmentation since the matched filter has been widely used in the detection of blood vessels of the human retina digital image.