Exudate detection in retinal fundus images using combination of mathematical morphology and Renyi entropy thresholding

Diabetic retinopathy (DR) is a microvascular complication of diabetes, causing abnormalities in the retina, and it is can cause blindness. Diabetic retinopathy can be detected by the appearance of hard exudates. Hard exudates are lipid formations leaking from these weakened blood vessels. Automatic detection of exudates is an early handler to diagnose diabetic retinopathy. This research proposed automatic detection of exudates using Renyi entropy thresholding and mathematical morphology. Renyi entropy thresholding has a controlling variable so that the obtained threshold value is more optimal. The proposed method using Renyi entropy thresholding and mathematical morphology has three stages: (1) preprocessing using contrast enhancement, (2) initial exudates detection based on mathematical morphology, and (3) exudates detection based on Renyi entropy thresholding. The test was performed using measurement evaluation method, sensitivity, specificity, and accuracy were 85.06%, 99.63%, and 99.54% respectively.