A precise image-based retinal blood vessel segmentation method using TAOD-CFNet

Yixin Yang, Lixiang Sun, Zhiwen Tang, Genhua Liu, Guoxiong Zhou, Lin Li, Weiwei Cai, Liujun Li, Lin Chen, Linan Hu

Abstract: Highlights•421 retinal vessel segmentation datasets from Zhuzhou Sany Aier Eye Hospital are precisely annotated.•We present the CFB to capture the difference between the sensitive and interfering regions of arteriovenous vessels.•We propose the ODGA to accurately localize and refine optimize the optic nerve disc region.•We design the TAM to enhance the edge features and context sensitive information of small and terminal blood vessels.