BOSlib-GPU 0.0.4

the library of Background Oriented Schlieren

BOSlib-GPU

This library is the extra pacage for BOSlib that makes GPU parallel processing available.

Key Features

• Short, concise code for 3D reconstruction and quantification
• GPU parallel processing

Warning

The BOSlib-GPU is still in its beta state. This means that it still might have some bugs and the API may change. However, testing and contributing is very welcome, especially if you can contribute with new algorithms and features.

Installing

1. Install PyTorch

Please install Pytorch 2.x . Make sure that CUDA is available on your PC. https://pytorch.org/get-started/locally/

2. Install torch_radon

Please install torch_radon　https://torch-radon.readthedocs.io/en/latest/getting_started/install.html

git clone https://github.com/matteo-ronchetti/torch-radon.git
cd torch-radon
python setup.py install

or

docker pull matteoronchetti/torch-radon

or if you are running Linux

wget -qO- https://raw.githubusercontent.com/matteo-ronchetti/torch-radon/master/auto_install.py  | python -

3. Install BOSlib

Use PyPI: https://pypi.org/project/BOSlib-GPU:

pip install BOSlib-GPU 

Or compile from source

Download the package from the Github: https://github.com/ogayuuki0202/BOSlib-GPU/archive/refs/heads/main.zip or clone using git

git clone https://github.com/ogayuuki0202/BOSlib-GPU.git
cd BOSlib-GPU
python setup.py install 

Methods

Please see our wiki below. Wiki

Contributors

Pleae refer original [BOSlib]((https://github.com/ogayuuki0202/BOSlib)