Geodesics in Shape Space via Variational Time Discretization
A variational approach to defining geodesics in the space of implicitly described shapes is introduced in this paper. The proposed framework is based on the time discretization of a geodesic path as a sequence of pairwise matching problems, which is strictly invariant with respect to rigid body motions and ensures a 1-1 property of the induced flow in shape space. For decreasing time step size, the proposed model leads to the minimization of the actual geodesic length, where the Hessian of the pairwise matching energy reflects the chosen Riemannian metric on the shape space. Considering shapes as boundary contours, the proposed shape metric is identical to a physical dissipation in a viscous fluid model of optimal transportation. If the pairwise shape correspondence is replaced by the volume of the shape mismatch as a penalty functional, for decreasing time step size one obtains an additional optical flow term controlling the transport of the shape by the underlying motion field. The implementation of the proposed approach is based on a level set representation of shapes, which allows topological transitions along the geodesic path. For the spatial discretization a finite element approximation is employed both for the pairwise deformations and for the level set representation. The numerical relaxation of the energy is performed via an efficient multi–scale procedure in space and time. Examples for 2D and 3D shapes underline the effectiveness and robustness of the proposed approach.
Unable to display preview. Download preview PDF.
- 2.Fuchs, M., Jüttler, B., Scherzer, O., Yang, H.: Shape metrics based on elastic deformations. Journal of Mathematical Imaging and Vision (to appear, 2009)Google Scholar
- 12.Eckstein, I., Pons, J., Tong, Y., Kuo, C., Desbrun, M.: Generalized surface flows for mesh processing. In: Eurographics Symposium on Geometry Processing (2007)Google Scholar
- 15.Miller, M.I., Younes, L.: Group actions, homeomorphisms and matching: a general framework. Technical report, John Hopkins University, Maryland (1999)Google Scholar
- 20.Kilian, M., Mitra, N.J., Pottmann, H.: Geometric modeling in shape space. ACM Transactions on Graphics 26(64), 1–8 (2007)Google Scholar
- 25.Black, M.J., Anandan, P.: A framework for the robust estimation of optical flow. In: Fourth International Conference on Computer Vision, ICCV 1993, pp. 231–236 (1993)Google Scholar
- 26.Kapur, T., Yezzi, L., Zöllei, L.: A variational framework for joint segmentation and registration. In: IEEE CVPR - MMBIA, pp. 44–51 (2001)Google Scholar