Paper Title
Localization Of Optic Disc In Retinal Images By Using An Efficient K-Means Clustering Algorithm

Abstract
In this paper, we proposed algorithm for localization of optic disc as it aids to detect different phases of diabetic retinopathy. The proposed algorithm gives accurate results for localization of optic disc. There are three stages defined in this paper for localization of optic disc. First stage is pre-processing in which the retinal image is enhanced. Second stage is clustering in which K-means algorithm is applied. Third stage is post-processing where morphological operations are performed. The proposed algorithm is applied on the DRIVE dataset. The results obtain when the number of clusters chosen are 16 and 20.