Open3D: A Modern Library for 3D Data Processing

Homepage | Docs | Quick Start | Compile | Python | C++ | Open3D-ML | Viewer | Contribute | Demo | Forum

Open3D is an open-source library that supports rapid development of software that deals with 3D data. The Open3D frontend exposes a set of carefully selected data structures and algorithms in both C++ and Python. The backend is highly optimized and is set up for parallelization. We welcome contributions from the open-source community.

Core features of Open3D include:

• 3D data structures
• 3D data processing algorithms
• Scene reconstruction
• Surface alignment
• 3D visualization
• Physically based rendering (PBR)
• 3D machine learning support with PyTorch and TensorFlow
• GPU acceleration for core 3D operations
• Available in C++ and Python

For more, please visit the Open3D documentation.

Installation on Jetson

Follow the instruction on Open3D ARM documentation to install the dependency first, for example:

sudo apt-get update -y
sudo apt-get install -y apt-utils build-essential git cmake
sudo apt-get install -y python3 python3-dev python3-pip
sudo apt-get install -y xorg-dev libglu1-mesa-dev
sudo apt-get install -y libblas-dev liblapack-dev liblapacke-dev
sudo apt-get install -y libsdl2-dev libc++-7-dev libc++abi-7-dev libxi-dev
sudo apt-get install -y clang-7

Then we need to change the default cblas.h:

cd /usr/include/aarch64-linux-gnu
sudo ln -sf cblas-netlib.h cblas.h

(optional)Then we can update the repository, by the method described by Github: update fork.

Then we clone the directory (shutong's copy)

git clone --recursive https://github.com/ricardodong/Open3D
cd Open3D
git submodule update --init --recursive
mkdir build
cd build

Then we do cmake, go to shutong_cmake.txt, copy and cmake command and run in terminal, for example:

cmake \
    -DCMAKE_BUILD_TYPE=Release \
    -DBUILD_SHARED_LIBS=ON \
    -DBUILD_CUDA_MODULE=ON \
    -DBUILD_GUI=ON \
    -DBUILD_TENSORFLOW_OPS=OFF \
    -DBUILD_PYTORCH_OPS=OFF \
    -DBUILD_UNIT_TESTS=ON \
    -DCMAKE_INSTALL_PREFIX=~/open3d_install \
    -DPYTHON_EXECUTABLE=$(which python) \
    ..

If Failed with CUDA compiler not found, do these first:

export CUDA_HOME=/usr/local/cuda
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/extras/CUPTI/lib64
export PATH=$PATH:$CUDA_HOME/bin

Then we can make: make -j$(nproc)

Then we can test cpp build:

make tests -j$(nproc)
./bin/tests --gtest_filter="-*Reduce*Sum*"

for the current build, we observe an error of not able to transform uint8 to float32:

C++ exception with description 
/home/correct-ai/Open3D/cpp/open3d/t/geometry/Image.cpp:163: 
Conversion from UInt8 to Float32 on device CUDA:0 is not implemented!

Then we do the installation (sudo not required!):

make install

At last we install python and we can test it:

make install-pip-package -j$(nproc)
python -c "import open3d; print(open3d)"

It's done!

Python quick start

Pre-built pip and conda packages support Ubuntu 18.04+, macOS 10.14+ and Windows 10 (64-bit) with Python 3.6-3.9.

# Install Open3D stable release with pip (including in conda virtual environments)
$ pip install open3d

# Test the installation
$ python3 -c "import open3d as o3d; print(o3d)"

To get the latest features in Open3D, install the development pip package. To compile Open3D from source, refer to compiling from source.

C++ quick start

Checkout the following links to get started with Open3D C++ API

• Compiling Open3D from source
• Open3D C++ API

To use Open3D in your C++ project, checkout the following examples

• Find Pre-Installed Open3D Package in CMake
• Use Open3D as a CMake External Project

Open3D-Viewer app

Open3D-Viewer is a standalone 3D viewer app available on Ubuntu and macOS. Please stay tuned for Windows. Download Open3D Viewer from the release page.

Open3D-ML

Open3D-ML is an extension of Open3D for 3D machine learning tasks. It builds on top of the Open3D core library and extends it with machine learning tools for 3D data processing. To try it out, install Open3D with PyTorch or TensorFlow and check out Open3D-ML.

Communication channels

• GitHub Issue: bug reports, feature requests, etc.
• Forum: discussion on the usage of Open3D.
• Discord Chat: online chats, discussions, and collaboration with other users and developers.

Citation

Please cite our work if you use Open3D.

@article{Zhou2018,
    author    = {Qian-Yi Zhou and Jaesik Park and Vladlen Koltun},
    title     = {{Open3D}: {A} Modern Library for {3D} Data Processing},
    journal   = {arXiv:1801.09847},
    year      = {2018},
}