Developability of Heightfields

Public code release for "Developability of Heightfields via Rank Minimization", presented at SIGGRAPH 2020 and authored by Silvia Sellán, Noam Aigerman and Alec Jacobson. Please note that while this Matlab implementation of our method is hereby released under MIT License, the method itself is pending a US patent filed in 2020 by Adobe Inc.

Installation

We have only succesfully validated this installation on MacOS 10.15 with Matlab 2020a, but are confident it should work on most Unix-based machines and versions of Matlab from the past five years.

To install this library, please start by cloning the repository recursively

git clone --recursive https://github.com/sgsellan/developability-of-heightfields.git

After this, we will build the mex functions in the gptoolbox directory:

cd developability-of-heightfields/gptoolbox/mex
mkdir build
cd build
cmake ..
make

Then, build our own mex files by entering Matlab and, in Matlab, adding this repository in its entirety to your Matlab path (for instance, by running addpath(genpath(path/to/developability-of-heightfields))), navigating to developability-of-heightfields/mex and running setup_mex in the Matlab console.

Use

To use our code in the exact same way we did for most our paper results, run in Matlab

gui_developables(path/to/some/mesh)

and follow the instructions in the console. For example, you can start by trying

gui_developables('data/bunny.obj')

We also allow you to replicate the results from our paper exactly, by runnning the scripts in the figures/ directory. Instructions to run and understand the output of each can be found in each scripts' first commented lines. In most cases, the scripts will generate input and output .obj files which we then rendered using Blender for the paper figures. You can look at our Blender setup in render/ and substitute the existing meshes with the input and output from our scripts to exactly replicate our paper figures up to very minor lighting direction and orientation choices.

C++ implementation

This Matlab implementation is the one we used to generate all the examples in the paper, and it has been thouroughly tested. I strongly encourage you to use it. However, if you do not have access to Maltab for whichever reason and running our code would be useful for you, I also provide a limited, guarantee-free C++ implementation of our algorithm here.

Known Issues

Please do not hesitate to contact sgsellan@cs.toronto.edu if you find any issues or bugs in this code, or you struggle to run it in any way.