/AlphaContours

A reference implementation of the Alpha Contours algorithm

Primary LanguageMATLABMIT LicenseMIT

This is a reference implementation of the 'Differential Operators on Sketches via Alpha Contours' by Mariia Myronova, William Neveu, and Mikhail Bessmeltsev, ACM Transactions on Graphics (SIGGRAPH 2023).

Structure of the repository:

AlphaContours/ -- C++ code of the main algorithm
SGP21_discreteOptimization-main/ -- a few files from https://github.com/llorz/SGP21_discreteOptimization with a minor change in their energy
algoB/ -- python code to compare with StrokeAggregator
paper/ -- Matlab scripts to reproduce figures in the paper
svg_to_matlab/ -- a Windows binary to convert .svg -> .m

Main algorithm

Command line: AlphaContours drawing.m [alpha]

Drawing.m is a vector drawing in a vector format (see the inputs folder). You can convert your own svg to .m using the binary in the 'svg_to_matlab' folder.

Alpha is the radius; when not specified, it is computed automatically.

The binary produces two files, all in Matlab format:

  • drawing_segments.m -- the segments added to the drawing
  • drawing_contours.m -- the Alpha Contours. This file contains three Matlab variables:
    • cppContour, a cell array containing exterior contours
    • cppContourInner, a cell array containing interior contours
    • seedPts, an array containing an arbitrary point inside each interior contour.

These variables can then be fed into Triangle library to triangulate the interior of the sketch shape, if necessary.

Dependencies

CGAL (tested with 5.5)

Eigen3 (tested with 3.4.0)

CMake

Building

We have only tried building under Windows, but the binary can be succefully run with wine under Mac. To build the C++ code

mkdir build
cd build
cmake ../

Then build the visual studio project as usual. Then move the binary, together with mpfr-6.dll and gmp-10.dll into AlphaContours/bin.

Reference Binary

If you don't want to compile anything, here is the reference binary for Windows. Unzip it into AlphaContours/bin. After that, you should be able to run reproduce all the figures by running the scripts in the paper/ folder, except for Fig.16 that requires installation of StrokeStrip and StrokeAggregator.

External code

To compute functional maps on our domains, we use the implementation of "Discrete Optimization for Shape Matching" by Jing Ren, Simone Melzi, Peter Wonka, and Maks Ovsjanikov from here. We have slightly modified their function compute_fmap.m.

To triangulate, we use the Triangle library by Jonathan Richard Shewchuk from here. It is 'triangle.exe' in our repository.

Citation

When using the code, please reference the following paper:

@article{AlphaContours,
author = {Myronova, Mariia and Neveu, William and Bessmeltsev, Mikhail},
title = {Differential Operators on Sketches via Alpha Contours},
year = {2023},
issue_date = {August 2023},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
volume = {42},
number = {4},
issn = {0730-0301},
url = {https://doi.org/10.1145/3592420},
doi = {10.1145/3592420},
journal = {ACM Trans. Graph.},
month = {jul},
articleno = {69},
numpages = {15},
keywords = {differential operators, vector graphics, sketch processing}
}