A class of fast geodesic shooting algorithms for template matching and its applications via the $N$-particle system of the Euler-Poincaré equations
For researchers in image registration and shape analysis, this work provides a computationally efficient method for template matching with theoretical convergence guarantees.
The paper introduces a class of fast geodesic shooting algorithms for template matching by reformulating the Euler-Poincaré equations as a finite-dimensional particle system, achieving up to O(N log N) computational cost via fast multipole methods. The algorithms demonstrate efficient feedback control iteration and accelerated convergence using conical kernels.
The Euler-Poincaré (EP) equations describe the geodesic motion on the diffeomorphism group. For template matching (template deformation), the Euler-Lagrangian equation, arising from minimizing an energy function, falls into the Euler-Poincaré theory and can be recast into the EP equations. By casting the EP equations in the Lagrangian (or characteristics) form, we formulate the equations as a finite dimensional particle system. The evolution of this particle system describes the geodesic motion of landmark points on a Riemann manifold. In this paper we present a class of novel algorithms that take advantage of the structure of the particle system to achieve a fast matching process between the reference and the target templates. The strong suit of the proposed algorithms includes (1) the efficient feedback control iteration, which allows one to find the initial velocity field for driving the deformation from the reference template to the target one, (2) the use of the conical kernel in the particle system, which limits the interaction between particles and thus accelerates the convergence, and (3) the availability of the implementation of fast-multipole method for solving the particle system, which could reduce the computational cost from $O(N^2)$ to $O(N\log N)$, where $N$ is the number of particles. The convergence properties of the proposed algorithms are analyzed. Finally, we present several examples for both exact and inexact matchings, and numerically analyze the iterative process to illustrate the efficiency and the robustness of the proposed algorithms.