Reinitialization common tools for algorithms.

Functions/Subroutines
subroutine	compute_curvature_filter (levelset, band, kappa_max, filter)
	This routine fills the curvature filter.

subroutine	compute_gradient_norm_error (phi, band, max_band_distance, error_l1l2linf, error_log)

subroutine	compute_centered_gradient_norm_error (phi, band, max_band_distance, error_l1, error_l2, error_linf)

subroutine	compute_gradient_norm_log_error (phi, band, max_band_distance, error_l1, error_l2, error_linf)

pure double precision function	compute_godunov_hamiltonian (gradp, gradm, phi)

pure double precision function, dimension(size(gradp))	compute_hamiltonian_gradient (gradm, gradp, phi)

pure double precision function	compute_ls_smooth_sign (phi, eps)

subroutine	smooth_field_outside_band (levelset, band, mindx)

Function/Subroutine Documentation

◆ compute_curvature_filter()

subroutine mod_lsm_reinitialization_tools::compute_curvature_filter	(	class(t_levelset), intent(in)	levelset,
		integer, dimension(:,:,:), intent(in)	band,
		double precision, intent(in)	kappa_max,
		double precision, dimension(:,:,:), intent(inout)	filter )

The curvature filter is scalar field such that:

filter(i,j,k) = 1 when the curvature is lower than kappa_max
filter(i,j,k) = 0 when the curvature is (much) higher than kappa_max
has an exponential decrease in between
Note
This filter is used in the HJ equation to avoid to deteriorate the levelset whenever the curvature is too high, ie. for very small structures

If kappa_max == 0 then the filter is 1