Main Article Content
Diabetic Retinopathy is an eye disease caused by Diabetes Mellitus. It is a major disease that has affected millions of people, the rate of people getting affected will increase exponentially in the upcoming years. It is caused by high blood sugar due to diabetes. It can be treated early during the starting stages, the vision problem will reduce. This project aims to create a system that can integrate image processing techniques together to predict whether the retinal image received from the patient is affected by DR or not. Color retinal images are important tools for diagnosing diabetic retinopathy. Images are captured from a fundus camera, stored on a computer, analyzed using a feed propagation neural network. Diabetic Retinopathy can be deducted by using Hemorrhages and Exudates deduction. After the elimination of the optic nerve and nerve cup, hemorrhages and exudates are deducted. The network is trained to recognize features in the retinal image. The digital filtering techniques and different features of extraction by GLCM are evaluated. ANN classification identifies the type of the diabetic retinopathy disease. The images are used to evaluate the neural network training tools for training state, regression, performance. The confusion matrix is also identified in this project.