ACVD

Info

This code is the implementation deriving from those papers:

[1] S. Valette,J.-M. Chassery and R. Prost, Generic remeshing of 3D triangular meshes with metric-dependent discrete Voronoi Diagrams, IEEE Transactions on Visualization and Computer Graphics, Volume 14, no. 2, pages 369-381, 2008.

[2] Sebastien Valette and Jean-Marc Chassery, Approximated Centroidal Voronoi Diagrams for Uniform Polygonal Mesh Coarsening, Computer Graphics Forum (Eurographics 2004 proceedings), Vol. 23, No. 3, September 2004, pp. 381-389.

[3] M. Audette, D. Rivière, M. Ewend, A. Enquobahrie, and S. Valette, "Approach-guided controlled resolution brain meshing for FE-based interactive neurosurgery simulation", Workshop on Mesh Processing in Medical Image Analysis, in conjunction with MICCAI 2011., Toronto, Canada, pp. 176--186, 09/2011.

This code is cross-platform and should compile under Linux, MacOS and Windows.

Licence

This code is distributed under the CeCILL-B license (BSD-compatible) (copyright CNRS, INSA-Lyon, UCBL, INSERM.)

Dependencies

• VTK www.vtk.org
• CMAKE www.cmake.org

Simple compilation HowTo under Linux

git clone https://github.com/valette/ACVD.git
cd ACVD
cmake . -DCMAKE_BUILD_TYPE=Release
make

the executables (ACVD, ACVDQ, AnisotropicRemeshingQ and others should be found under the "bin" subdirectory)

Usage

execute ACVD and ACVDQ without arguments to see the available options.

when using graphical display, the 'e' key allows to continue to the next step during interaction

for ACVD, the output is written in the file simplification.ply

additionnally, when running ACVD, a file output_1.ply is also written. It is the output mesh before post-processing using quadrics.

note that to enforce a manifold output mesh, such as explained in [3], you need to use the -m 1 option.

comments, suggestions : https://github.com/valette/ACVD/issues

Output mesh quality vs complexity

ACVD works by clustering input vertices. The accuracy and output mesh quality is therefore directly dependent on two important factors:

• the ratio between the number of input vertices and the number of output vertices. If this ratio is bellow a predefined value, the input mesh will be subdivided until the ratio is satisfactory. By default, the ratio is 10. Increasing this value (using the -s ratio) parameter greatly improves the output mesh quality. Example : "-s 100"

• the input mesh quality. If the input mesh has elongated triangles, artifacts in the output mesh can appear. The parameter "-l ratio" is here to solve this problem. If this parameter is set, ACVD will split all input mesh edges that are longer than ratio*average_edge_length. Example : "-l 3"

Multithread versions

For each program ACVD, ACVDQ and AnisotropicRemeshingQ, there is a parallel implementation, called ACVDP, ACVDQP and AnisotropicRemeshingQP. In the examples bellow, just add a trailing "P" to the executable to use all your processor cores. Note that the parallel versions are not deterministic, so running the programm twice with the same parameters will yield different remeshings. The parallel versions run much faster when quadrics are used (i.e. with ACVDQ or AnisotropcRemeshigQ), but the speedup is small with linear ACVD. For all programs, the number of threads can be set using the "-p numberOfThreads" option.

Examples

Remeshing the Stanford bunny to 3000 vertices :

wget https://github.com/alecjacobson/common-3d-test-models/raw/master/data/stanford-bunny.obj
bin/ACVD stanford-bunny.obj 3000 0

taking into account curvature:

bin/ACVD stanford-bunny.obj 3000 1.5

Remeshing the Stanford bunny to 3000 vertices with high output mesh quality:

bin/ACVD stanford-bunny.obj 3000 0 -s 100 -l 3

Remeshing the fandisk to 3000 vertices, taking into account sharp features with ACVDQ:

wget https://github.com/alecjacobson/common-3d-test-models/raw/master/data/fandisk.obj
bin/ACVDQ fandisk.obj 3000 0

Remeshing the horse to 1000 vertices with anisotropic metric:

wget https://github.com/alecjacobson/common-3d-test-models/raw/master/data/horse.obj
bin/AnisotropicRemeshingQ horse.obj 1000 1.5

Remeshing the Thai Statue to 100000 vertices with curvature computation

wget http://graphics.stanford.edu/data/3Dscanrep/xyzrgb/xyzrgb_statuette.ply.gz
gunzip xyzrgb_statuette.ply.gz
bin/ACVDQ xyzrgb_statuette.ply 100000 1.5

parallel version:
bin/ACVDQP xyzrgb_statuette.ply 100000 1.5

parallel version restricted to 3 threads:
bin/ACVDQP xyzrgb_statuette.ply 100000 1.5 -np 3

for all the examples above, interactive visualization of the processing can be triggered by adding "-d 2" to the command lines

Python port

A part of ACVD has been ported to python here: https://github.com/pyvista/pyacvd