The FAC model performs slightly better than the CV model and the GACV model in terms of [P.sub.mp].
Figures 10(b), 10(c), and 10(e) show that the CV model, the GACV model, and the FAC model fail to obtain the right segmentation results, as the top dot cannot be detected.
Two paws cannot be detected by the GACV model as shown in Figure 13(c).
Figures 14(b), 14(c), and 14(e) depict that the CV model, the GACV model, and the FAC model are unable to correctly detect the real contour of all the rice grains, especially the rice grains at the bottom of the image.
Figure 15 shows that the results obtained by the GACV model and the FAC model cannot detect the whole vessel, as the vessel at the bottom left of the image is submerged into the background.
Segmentation results obtained by the CV model, the GACV model, and the DRLSE model are sensitive to different initial conditions.
Yang, "GACV: geodesic-aided C-V method," Pattern Recognition, vol.