摘要： Diabetic Retinopathy (DR) is an eye abnormality in which the human retina will get affected and is becoming one of the leading cause of preventable blindness. In the world, it is found that nearly 4.8% of blindness is caused due to DR. Preliminary symptoms include the formation of microaneurysms, exudates and hemorrhages. Early detection of DR can save the vision of diabetes patients and manual diagnosis takes time and effort for confirmation. In this paper, a Computer-aided Automated Diagnosis (CAD) is developed to solve this problem. The proposed approach uses edge-based segmentation method for segmenting the optic disc and blood vessels more accurately than region-based methods, followed by extraction of most probable exudates regions, feature extraction and the classifier stage to detect the presence of exudates. This system achieved sensitivity 82.61%, specificity 92.31% and moreover an accuracy of 87.75% for DIARETDB dataset.