Diabetic Retinopathy Detection using RGB Image Processing Techniques

Abstract

In modern era diabetic eye disease comprises a group of eye conditions that affect people with diabetes, in which diabetic retinopathy (DR) has the potential to cause severe vision loss. This paper proposes a peculiar algorithm in MATLAB for automating the detection of DR using fundus image. Several morphological operations performed on fundus image to detect and extract the features like blood vessels, micro-aneurysms and exudates. Automatic detection of DR will determine to sort between normal or abnormal eye based on the features detection. As a result, an early detection of DR facilitates medication to prevent vision loss. Real-time image processing shows encouraging results in terms of validity and efficiency.