Efficient Detection of severity of Diabetic Retinopathy using DIP

Diabetes is a rapidly increasing worldwide problem which is characterized by defective metabolism of glucose that causes long-term disorders and failure of various organs of human body. Diabetic Retinopathy (DR) is a most common complication of diabetes. Nowadays it is the primary causes of blindness and visual impairment in adults. This can be prevented if diagnosed and treated in its early stages by detecting the microaneurysms(MAs) and Exudates in the retina of the diabetic patient. The proposed work develops an efficient system for ophthalmologist to analyze the MAs and exudates. The abnormalities in the captured color fundus image is detected using digital image processing(DIP)techniques by applying morphological and other signal processing operations. The extracted features are used to detect the severity of DR with accuracy of 94%.