Automatic Detection of Optic Disc for Diabetic Retinopathy

Abstract

Diabetic Retinopathy affects the retina of the eye. Ultimately it may lead to total visual impairment. Total blindness can be avoided by detecting Diabetic Retinopathy at an early stage. Various manual tests are used by the doctors to detect the presence of disease but they are tedious and expensive. Some of the features of Diabetic Retinopathy are exudates, haemorrhages and micro aneurysms. Detection and removal of optic disc plays a vital role in extraction of these features. This paper focuses on detection of optic disc using various image processing techniques, algorithms such as Canny edge, Circular Hough(CHT). Retinal images from IDRiD, Diaret_db0, Diaret_db1, Chasedb and Messidor datasets were used.