Skip to main content
Figure 9 | EURASIP Journal on Image and Video Processing

Figure 9

From: Structure-based level set method for automatic retinal vasculature segmentation

Figure 9

Level set evolution. Synthetic image (a), initialization of the level set function with a binary function using c0 = 5 (b), and its initial zero-level contour (c), image edges from modified phase map (d), final level set contours based on the proposed method using the potential function P1(e), and its level set function (f), final level set contours based on the proposed method using the potential function P2(g), and its level set function after 451 iterations (h), slope of final level set function in a band region with size of 2c0(i), final zero-level contours of the given image based on DRLSE[25] using the potential function P2 and set by a negative-valued α (j) and a positive-valued α (k). In (g), with area functional A(.), the initial contour is shrunk and expanded automatically to match the boundary of vessels. With the length functional L(.), this fitting has become smooth. The initial level set function in (b) is regularized using the regularization functional R(.), and the final, regularized level set function in (h) is obtained.

Back to article page