Amna Iftikhar


Master of Science in Computer Science

Faculty / School

Faculty of Computer Sciences (FCS)


Department of Computer Science

Date of Submission



Dr. Muhammad Saeed

Project Type

MSCS Survey Report


Diabetic retinopathy affects the vision of diabetic patient due to the damage of tiny blood vessels at the back of retina. If left undiagnosed and untreated, diabetic retinopathy can result initially in mild vision problems and then, to complete blindness. Timely detection followed by suitable treatment can help many people from going through the agony of vision loss. Image processing and computer vision techniques can be used for the automatic detection of Diabetic Retinopathy. This report discusses an effective algorithm to detect and localize optical disc, blood vessels and exudates in retinal fundus images. The algorithm applies various computer vision and morphological image processing techniques to diagnose and distinguish exudates in diabetic retinopathy patients. The proposed algorithm is tested on five retinal images that were hand marked by an ophthalmologist. We obtained 82.1% predictive value and 88.9% sensitivity with our algorithm.

The full text of this document is only accessible to authorized users.