Abstract
Glaucoma is the world's first irreversible blindness-causing ophthalmic disease, and early screening and treatment of glaucoma are crucial for patients. However, glaucoma patients have no obvious symptoms in the early stage of the disease, and manual screening would be labor-intensive, so there is an urgent need for an automated system for primary glaucoma screening. Recently, great expansion and pledge have been displayed in the analysis of eyeground images through applying deep learning methods. In the existing research system, there are many machine learning methods with image recognition techniques are used for segmentation of optic cup optic disc region, and in order to improve the recognition accuracy, the existing diagnostic techniques need to be improved. In this paper, we introduce the contents related to fundus image based glaucoma diagnosis, aiming to help readers understand the importance of optic cup optic disc region in glaucoma diagnosing and some existing deep learning algorithms in respect of optic cup optic disc region partition at home and abroad, and list the research gaps and challenges for the purpose of enhancing the precision of segmentation methods.