DETECTION AND CLASSIFICATION OF DIABETIC RETINOPATHY DISEASES :A SURVEY

  • Atika Rahman

Abstract

With the increase in the number of diabetic patients all over the world, a large number of populations are suffering from eye diseases. Diabetic retinopathy (DR) is an important cause of blindness in the world today. However, DR is hard to detect and diagnose and the procedure of diagnosis is time-consuming. The computer-aided diagnosis based on deep machine learning algorithms is proposed in this project. The project addresses the detection of hemorrhages and microaneurysms in color fundus images. A neural network is trained with a transfer learning technique and the classification is done using the SVM classifier. Different from the previous works, the proposed project works on Pre-processing of the fundus images of the retina and extract the GLCM(Gray Level Co-Occurrence Matrix) feature, after which the classifier will predict about the retinal diseases whether hemorrhages or microaneurysms. After predicting the localization, the segmentation would be able to detect the disease with more accuracy.

Published
2019-12-11
Section
Articles