songsenand
/
cgal
mirror of https://github.com/CGAL/cgal
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Revert "Follow the OG paper's prescription of the SVD threshold" 

This reverts commit 913f8796c9.

While a larger epsilon bound gave better results for non-manifold
pinched configurations, it also resulted in a lot of sharpness
loss.

For now, I'd rather it does not do well at non-manifold configurations.

Improving results on those configurations is todo.
This commit is contained in:
Mael Rouxel-Labbé 2024-02-27 15:57:49 +01:00
parent ed1e4fb806
commit b9411822a4
1 changed files with 2 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
Isosurfacing_3/include/CGAL/Isosurfacing_3/internal/dual_contouring_functors.h
View File

@ -168,9 +168,8 @@ bool cell_position_QEM(const typename Domain::Cell_descriptor& c,
Eigen::JacobiSVD<typename Eigen_matrix_x::EigenType> svd(A, Eigen::ComputeThinU | Eigen::ComputeThinV);
// Lindstrom's paper, "Out-of-Core Simplification of Large Polygonal Models": 1e-3
// Ju's paper, "Dual Contouring of Hermite Data": 1e-1
svd.setThreshold(1e-1);
// set threshold as in Peter Lindstrom's paper, "Out-of-Core Simplification of Large Polygonal Models"
svd.setThreshold(1e-3);
Eigen_vector_3 x_hat;
x_hat << x_coord(com), y_coord(com), z_coord(com);