Title :
Simulated static electric field (SSEF) snake for deformable models
Author :
Yuan, Dan ; Lu, Siwei
Author_Institution :
Dept. of Comput. Sci., Memorial Univ. of Newfoundland, St. John´´s, Nfld., Canada
Abstract :
In this paper a novel design of external force for snake is proposed. This kind of external force is actually a field, which we call the simulated static electric field. This field is created by simulated static electric charges, and the forces in this field are computed by the Coulomb law which is used in the analysis of static charge fields. The existing problems, such as the capture range, initialization, noise sensitivity and poor convergence to concavities, which appeared in the other designs of snake can be largely solved by this new scheme. Also, cracked boundaries extracted from real images can be located. Several comparisons with gradient vector flow snake are presented in this paper to show the performance of our snake.
Keywords :
edge detection; electric fields; feature extraction; image processing; image segmentation; Coulomb law; active contours; capture range; cracked boundary extraction; gradient vector flow snake; initialization; noise sensitivity; parametric deformable models; simulated static electric field; Active contours; Active noise reduction; Analytical models; Computational modeling; Computer science; Computer vision; Convergence; Deformable models; Equations; Image converters;
Conference_Titel :
Pattern Recognition, 2002. Proceedings. 16th International Conference on
Print_ISBN :
0-7695-1695-X
DOI :
10.1109/ICPR.2002.1044618