A Linear Elastic Force Optimization Model for Shape Matching

We employ an elasticity based model to account for shape changes. In general, to solve the underlying equations for the deformation, boundary conditions have to be incorporated, e.g., in the form of correspondences between contour points. However, exact boundary correspondences are usually unknown. We propose a method that is able to optimize pre-selected boundary conditions such that external forces causing the shape change are minimized in some sense. Thus we seek simple physical explanations of shape change close to a pre-selected deformation. Our method decomposes the full nonlinear optimization problem into a sequence of convex optimizations. Artificial and natural examples of shape change are given to demonstrate the plausibility of the algorithm.