Color fundus image blood vessel segmentation method based on U-net convolutional neural network

The invention discloses a color fundus image blood vessel segmentation method based on a U-net convolutional neural network, belongs to the technical field of image processing and deep learning, and particularly relates to a fundus image blood vessel automatic segmentation method based on a neural network. According to the method, Gamma correction is carried out after contrast-limited adaptive histogram equalization processing is carried out on the image, so that the contrast of blood vessels in the fundus image and the background of the fundus image can be effectively improved, the filteringeffect of the two-dimensional Gabor filter is improved, and segmentation of micro blood vessels in the fundus image is facilitated. According to the invention, after the two-dimensional filter, the image is subjected to the minimization processing, so that the noise can be well suppressed while the blood vessel is segmented. The adopted neural network is an improved U-net neural network, the sizeof an input image is 48 *