Automatic location of Optic Disc boundary in retinal images

The localization of Optic Disc (OD) is an initial step for the segmentation of various pathological features like microaneurysms, hemorrhages, exudates etc. We propose a new technique for the localization of OD boundary in retinal images. The retinal image consists of blood vessels that emerge from the OD. The presence of these blood vessels may act as a disturbance for the localization of OD. Cumulative histogram is applied on the morphologically processed green band of the retinal image to segment the blood vessels. Then the blood vessels are removed from the red band of the image where the Optic Disc is clearly seen. Further the removal of blood vessel is done using a method say anisotropic diffusion. Maximum intensity pixel is assumed as a centroid of the OD and the gradient for each degree is calculated. The local maximum points are identified from the gradient. The local maximum point closest to the center of the Optic Disc is chosen as the contour points. The results are compared with a ground truth image from the ophthalmologist. The source retinal image for performing this work is obtained from the publicly available STARE database. This method offers a successful localization of OD which may help in diagnosis of various retinal abnormalities.