Optic Disc Segmentation in Retinal Fundus Images Using Morphological Techniques and Intensity Thresholding

Abstract

Among the many diseases that affect the retina, there are two serious diseases: Diabetic Retinopathy and Glaucoma. Diabetic Retinopathy (DR) is a sight-threatening risk inflicting disorder diabetic patient. It occurs due to damage in the retina as a result of diabetes. Glaucoma is a disease of the retinal system that damages the optic nerve of the eye and gets worse over time. A buildup of pressure inside the eye is often associated with it, so it can damage the optic nerve that transmits images to the brain. If damage to the optic nerve caused by high eye pressure continues, glaucoma causes permanent vision loss. Early diagnosis and treatment have been shown to prevent blindness and visual loss. Compared with the manual diagnostic methods, automated retinal analysis systems help save patients’ time, cost and vision. In this context, a fundamental process for diagnosing these diseases is the precise and effective localization of the Optic Disc (OD) in retinal images. This paper offers a robust and efficient method for automatic diagnosis of OD. The method starts with removing the undesirable portions of the image, such as noise, reflections and blur. Next, the OD has been detected by means of morphological operations considering the intensity threshold feature. The proposed method is fast and robustness to detect OD even if interrupted by the visible blood vessels. This method was applied on seven data sets which are the Origa, Rim-One 3, Drishti, Messidor, Drions, Diaretdb0 and DIARETDBI, and the resulted accuracy for these data set are 98.46, 97.48, 97.03, 98.75, 97.27, 95.38, 95.45 respectively.