Skip to content

Fast code for parallel or fan beam tomographic reconstruction

License

Notifications You must be signed in to change notification settings

cabouman/svmbir

Repository files navigation

svmbir

Python code for fast MBIR (Model Based Iterative Reconstruction)
This is a python wrapper for High Performance Imaging's supervoxel C code, HPImaging/sv-mbirct.

Full documentation is available at svmbir_docs.

To cite this software package, please use the bibtext entry at cite_svmbir.

Installing svmbir

Currently supporting Python 3.9-3.12, on MacOS and Linux (Windows possible but not actively maintained).

svmbir packages are available from conda-forge and PyPI, or can be built and installed from source.

  • (recommended) Create a clean virtural environment, such as
conda create -n svmbir python=3.10
conda activate svmbir
  • To install from conda-forge,
conda install -c conda-forge svmbir
  • To install from PyPI,
pip install svmbir
  • Installing from source (requires GNU/gcc compiler, OMP libraries),
# In top repository folder,
CC=gcc pip install .        # also supports Intel "icc"

See here for more details.

Running the demos

  1. Download demo.zip at https://github.com/cabouman/svmbir/blob/master/demo.zip.
  2. Uncompress the zip file and change into demo folder.
  3. In your terminal window, install required dependencies of demo.
pip install -r requirements_demo.txt
  1. In your terminal window, use python to run each demo.