Figure 1From: Structure-based level set method for automatic retinal vasculature segmentationSample retinal fundus images and manual segmentations. A pathological image (of size 640 × 480 pixels) from our own dataset ([2, 19]) given in (a), a non-pathological image (of size 565 × 584 pixels) from the DRIVE dataset ([3]) given in (b), the manual segmentation image of (a) given in (c), and the manual segmentation image of (b) given in (d).Back to article page