Eye fundus image blood vessel segmentation method and system

The invention discloses a fundus image blood vessel segmentation method and system. The method comprises the following steps: collecting a data set of a fundus image; preprocessing the collected data set to obtain a prepared data set; constructing a fundus image blood vessel segmentation model according to the preparation data set; by training a fundus image blood vessel segmentation model and verifying the accuracy of the model, the accuracy of blood vessel detail segmentation is improved. The retinal blood vessel segmentation task is deeply studied, aiming at the fundus image blood vessel segmentation problem, starting from image preprocessing and a segmentation model based on deep learning, different image enhancement technologies are adopted, and a blood vessel segmentation network is improved through feature analysis of fundus image blood vessels; the invention further provides a retinal vessel segmentation method based on the GAN. According to the method, automatic segmentation of the blood vessels in t