Fig. 6

Visualization of the retinal segmentation process, showcasing the original full retinal image, the corresponding full ground truth, the extracted retinal patches, their ground truths, and the segmentation result produced by the proposed UGS-M3F method on the STARE dataset