C++ tools for 3D data

Toolbox of low level functions for 3D shape manipulation. The data structure is similar to the one used in libIGL (Eigen::MatrixXd for vertices and Eigen::MatrixXi for faces), but prioritise XYZ to be stored in the rows.

Content

IO

• read / write PLY files (tinyply)
• read / write OFF files (libigl)
• read / write OBJ files (libigl)
• read / write STL files (libigl)
• read / write CSV files (stackoverflow)
• write PNG file (stb)
• return list of files in folder (BOOST)

Convertors

• IO to CGAL

EigenTools

• wrapper for nanoflann
• remove cols /rows
• standardDeviation(Eigen::VectorXd)
• variance(Eigen::VectorXd)
• concatenate 2 matrices (sparse / dense)
• generate sparse diagonal matrix

STDTools

• return the sign

mesh

• compute the volume and surface area of a closed 3D mesh
• compute the normals from a mesh
• compute the faces centroids
• knn search on the manifold (greedySearch class)
• fill the holes (require CGAL)
• mesh downsampling (require CGAL)

clouds

• centering to (0, 0, 0)
• cloud downsampling (skip points)
• Poisson surface reconstruction (require CGAL)

visualization

• mesh / two meshes
• cloud
• correspondences

Dependencies

Standard dependencies

sudo apt-get install libeigen3-dev libxrandr-dev libxinerama-dev libxcursor-dev libxi-dev libboost-all-dev libyaml-cpp-dev

Some of the algorithms require CGAL which can be installed as:

sudo apt-get install libcgal-dev

To use

Clone the git repository and generate a minimalistic project:

git clone https://github.com/rFalque/3D_tools.git
cd utils
sudo chmod +x sample/setup.sh
./sample/setup.sh

In the CMakeLists.txt of the main project add

add_subdirectory(utils)

...

target_link_libraries(...
                      utils
                      ...
)

formating

auto formatting is used with clang-format using:

find ./ -iname *.h -o -iname *.cpp | xargs clang-format -i -style=Google