When having a very complex structure to represent (many layers and faults, triple points and interpolation data), the authors have to find a new method which takes into account the particular aspect of the data. The originality of this segmentation method is to consider the active contour model as a set of articulated curves, which corresponds to the interfaces between different regions taking into account "triple points". Moreover, the a priori knowledge of interpolation data allows one to make some geometric constraints on the model. The solution is obtained by minimization of a nonlinear functional under constraints in a suitable convex set. The geometrical constraints are associated with interpolation data. Deformable models allow to interactively act on the representation on adding a dynamic term in the minimization problem that allows to upgrade the models to the solution of the minimization problem introduced in the modelization. In regards to the usual deformable techniques, the characteristics of the proposed method are: the use of several potentials to treat only one object, a k-order Taylor series on time of the nonlinear terms linked to the potentials to take into account much more voxels as in other methods, the use of the interpolation data on the objects, which have the same consequence on the model as the "balloons" forces. Numerical results on real data are given.
[1]
Josef Kittler,et al.
Edge-Labeling Using Dictionary-Based Relaxation
,
1990,
IEEE Trans. Pattern Anal. Mach. Intell..
[2]
ISAAC COHEN,et al.
Using deformable surfaces to segment 3-D images and infer differential structures
,
1992,
CVGIP Image Underst..
[3]
Guillermo Sapiro,et al.
Geodesic Active Contours
,
1995,
International Journal of Computer Vision.
[4]
Laurent D. Cohen,et al.
Deformable models for 3-D medical images using finite elements and balloons
,
1992,
Proceedings 1992 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.
[5]
Laurent D. Cohen,et al.
Surface reconstruction using active contour models
,
1993
.
[6]
James A. Sethian,et al.
A unified approach to noise removal, image enhancement, and shape recovery
,
1996,
IEEE Trans. Image Process..
[7]
Laurent D. Cohen,et al.
On active contour models and balloons
,
1991,
CVGIP Image Underst..