A header-only C++ library with full Python bindings for handling, converting, and storing 3D motions and poses.
It supports converting, chaining and applying transformations in many common representations:
- Axis-angle rotation and translation vector
- Dual quaternion
- Euler angles and translation vector
- SE(3) transformation matrix
- Rotation quaternion and translation vector
Using the TransformInterface
base class, transformations can be applied and chained without worrying about the underlying type.
The TransformContainer
enables handling multiple transformations as motions or as poses, with or without corresponding timestamps.
Furthermore, the M3DWriter
and M3DReader
can be used to store and load transformations from both C++ and Python in the M3D file format.
Supported operating systems:
- Ubuntu 20.04
- Ubuntu 22.04
Dependency | Version | Note |
---|---|---|
CMake | >= 3.14 | |
GCC | >= 9.4 | C++17 is required |
Eigen | >= 3.3 | |
Python | >= 3.8 | Tests are run for 3.8, 3.10, and 3.12 |
All dependencies can be installed using:
apt install build-essential cmake libeigen3-dev python3 python3-dev python3-pip
See docker/motion3d.dockerfile
for further installation details, e.g., for packages required for development.
After all requirements are available, use CMake and Make to install the library headers:
mkdir build && cd build
cmake ..
make install
For uninstalling use make uninstall
.
The library can be installed either over PyPI using pip:
python3 -m pip install motion3d
or directly from within the cloned repository:
python3 -m pip install .
If Eigen is not found on the system, the setup downloads a suitable version locally before building the package.
Fore more information about the transformation types, definitions, and the usage in C++ and Python, see the documentation at motion3d.readthedocs.io.
Furthermore, the doc
directory provides a detailed readme on how to create the documentation directly from the repository.