Segmentation of optic disc and cup in fundus images using maximally stable extremal regions

In procedure of clinical routine on retinal fundus, Optic Disc and Cup are often used for diagnosis many diseases such as Glaucoma, Diabetic retinopathy and Age-Related Macular Degeneration (AMD). Optic Disc and Cup are the central organ of retinal fundus which control the balance of blood vessels. Medical practitioners often used different information on these organs to diagnose such diseases. In this work, we proposed a method to automatically segment the Optic Disc and Cup from a low contrast retinal fundus image for further diagnosis by the doctor. The proposed method consists of three main steps including pre-processing, region localization and Optic Disc and Cup extraction. Firstly, Red-Channel and Morphological techniques are applied as a preprocessing step. Red-channel Method is used to enhance properties of the disc area. Then, Morphological Technique is used to remove blood vessels from the retinal fundus. Secondly, Region Localization Method using Maximal Stable Extremal Region is used to Optic Disc and Cup. Finally, Optic Disc and Cup are extracted from the retinal fundus image to calculate the disk to cup ratio.