juliageometry / marchingcubes.jl Goto Github PK

View Code? Open in Web Editor NEW

18.0 3.0 4.0 745 KB

Efficient Implementation of Marching Cubes' Cases with Topological Guarantees

License: MIT License

Julia 46.89% C 53.11%

marching-cubes isosurface-extraction implicit-surface-tiler topological-guarantees

marchingcubes.jl's Introduction

MarchingCubes

Julia port of Efficient Implementation of Marching Cubes' Cases with Topological Guarantees.

Public article available here.

Implementation

Adapted to Julia (1 based indexing) from the original c++ implementation (0 based indexing).

Tiny benchmark - visualization with ParaView

$ g++ -O3 ply.c main.cpp MarchingCubes.cpp
$ ./a.out
Marching Cubes ran in 0.007834 secs.
$ julia --check-bounds=no
julia> using BenchmarkTools, MarchingCubes
julia> mc = MarchingCubes.scenario();
julia> @btime march($mc)
  7.865 ms (0 allocations: 0 bytes)
julia> @btime march_legacy($mc)
  9.268 ms (0 allocations: 0 bytes)
julia> using PlyIO
julia> MarchingCubes.output(PlyIO, mc)  # writes "test.ply" (can be opened in a viewer, e.g. ParaView)

Test scenario output:

MWE demonstrating visualization with MeshViz.jl

using MarchingCubes
using GLMakie
using MeshViz
using Meshes

mc = MarchingCubes.scenario()
march(mc)
msh = MarchingCubes.makemesh(Meshes, mc)
display(viz(msh))

MWE demonstrating visualization with GeometryBasics.jl and Makie.jl

using GeometryBasics
using MarchingCubes
using GLMakie

mc = MarchingCubes.scenario()
march(mc)
msh = MarchingCubes.makemesh(GeometryBasics, mc)

fap = mesh(msh; color = :gray)
display(fap)

Original BibTeX

@article{marching_cubes_jgt,
    author = {Thomas Lewiner and Hélio Lopes and Antônio Wilson Vieira and Geovan Tavares},
    title = {Efficient implementation of marching cubes cases with topological guarantees},
    year = {2003},
    month = {december},
    journal = {Journal of Graphics Tools},
    volume = {8},
    number = {2},
    pages = {1--15},
    publisher = {A.K.Peters},
    doi = {10.1080/10867651.2003.10487582},
    url = {\url{http://thomas.lewiner.org/pdfs/marching_cubes_jgt.pdf}}
}

License

This code is free to use under the terms of the MIT license.

marchingcubes.jl's People

Contributors

Stargazers

Watchers

Forkers

t-bltg ickaser abbooy nickkeepfer

marchingcubes.jl's Issues

mesh simplification example

Provide an example of mesh simplification using https://github.com/JuliaGeometry/QuadricMeshSimplification.jl.

Proposed in #21 (comment).

faces with zero volume in large models

The following is a mesh analysis with SpaceClaim of a mesh generated in C with the LewinerMarchingCubes from NeurolabUSC, modified to accept Int64 and thus large numbers of voxels and arrays instead of NIIFTI:

The listed 'errors' are not actually errors.

However, with MarchingCubes.jl, there are 8514 faces with zero volume, out of 5,421,474 faces:

I suspect that the 2000 x 2000 x 1995 voxels are hitting an Int32 limit somewhere, or there is a slight error in the LUT's due to the 1-index conversion. I think the latter is more plausible. The source data is 30 GB, so we'd need to arrange some way to transfer this if you are interested in trying to reproduce the result.

Feature request: make mesh more amenable to other Julia datatypes?

One of the reasons I want to use this module is to be able to put together a pure-Julia data pipeline for micro-CT data. To that end, it would be helpful if I could take the mesh generated by this project, and pull it over to Makie.jl or some other Julia plotting software.

It seems to me like it shouldn't be too difficult to modify the code base to use Triangles from GeometryBasics.jl , then make the mesh compatible with Meshes.jl and therefore MeshViz .

I can help with this, I think (although maybe not very fast).

"Invalid triangles" when isosurface value is not 0.0

Something goes wrong with the algorithm when I use a nonzero isosurface value.

Here's my MWE:

using MarchingCubes
xi = [1.0, 2, 3, 4, 5]
dat = [(x-3)^2 + (y-3)^2 + (z-3)^2 for x in xi, y in xi, z in xi]
datsub = dat .- 5.0

m1 = MC(dat)
march(m1, 5.0)

m2 = MC(datsub)
march(m2)

march(m1, 5,0) gives me a lot of warnings about invalid triangles:

┌ Warning: Invalid triangle 6
└ @ MarchingCubes C:\Users\iwheeler\.julia\packages\MarchingCubes\V0hg7\src\MarchingCubes.jl:557
┌ Warning: Invalid triangle 12
└ @ MarchingCubes C:\Users\iwheeler\.julia\packages\MarchingCubes\V0hg7\src\MarchingCubes.jl:557

on up to triangle 128, with many repeats. On the other hand,
march(m2) runs just fine, and when I examine its output with PlyIO and Paraview, it gives me the sphere I expect.

TagBot trigger issue

This issue is used to trigger TagBot; feel free to unsubscribe.

If you haven't already, you should update your TagBot.yml to include issue comment triggers.
Please see this post on Discourse for instructions and more details.

If you'd like for me to do this for you, comment TagBot fix on this issue.
I'll open a PR within a few hours, please be patient!

Should the output function only be in the example file?

It seems kind of fragile to not have the output function in the main file, MarchingCubes.jl.