1: \begin{abstract}

2: We study parameterisation-independent closed planar curve matching

3: as a Bayesian inverse problem. The motion of the curve is modelled 

4: via a curve on the diffeomorphism group acting on the ambient space,

5: leading to a \emph{large deformation diffeomorphic metric mapping} 

6: (LDDMM) functional penalising the kinetic energy of the deformation.

7: We solve Hamilton's equations for the curve matching problem using

8: the Wu-Xu element  [S. Wu, J. Xu, Nonconforming finite element spaces

9: for $2m^{\text{th}}$ order partial differential equations on $\mathbb{R}^n$ simplicial 

10: grids when $m = n + 1$, Mathematics of Computation 88 (316) (2019)

11:  531–551] which provides 

12: mesh-independent Lipschitz constants for the forward motion of the

13: curve, and solve the inverse problem for the momentum using Bayesian

14: inversion.  Since this element is not affine-equivalent

15: we provide a \emph{pullback theory} which expedites the implementation

16: and efficiency of the forward map. We adopt ensemble Kalman inversion using

17: a negative Sobolev norm mismatch penalty to measure the discrepancy

18: between the target and the ensemble mean shape. We provide

19: several numerical examples to validate the approach.

20: \end{abstract}

21: