OCTA fundus blood vessel extraction and quantification method based on region connectivity method

The invention discloses an OCTA fundus blood vessel extraction and quantification method based on a region connectivity method, and the method comprises the steps: firstly converting an input image into a gray-scale image, and weakening noise points in the image through image processing methods such as smoothing; then performing detection operation of blood vessel starting and connected regions on the pre-denoised image from different angles, combining blood vessel detection results from different angles to obtain a complete blood vessel network image, performing traversal smoothing, and removing near-point discrete noise suspended around the blood vessel; then, carrying out binarization processing and corrosion operation on the blood vessel image to respectively obtain a binary image and a skeleton image of the blood vessel; finally, data processing is carried out, the total area and length of the blood vessels are solved through a pixel accumulation method, and the average width of the blood vessels is obta