This code reproduces all the results presented in Core Imaging Library part II: multichannel reconstruction for dynamic and spectral tomography.

Instructions

1) Install the environment

Note: Depending on your nvidia-drivers, you can modify the cudatoolkit parameter. See here for more information.

conda create --name cil2_demos -c conda-forge -c astra-toolbox/label/dev -c ccpi cil cil-astra ccpi-regulariser nb_conda_kernels jupyterlab scikit-image python-wget cudatoolkit=_._

2) Activate the environment

conda activate cil2_demos

Create the environment from the requirements.yml file:

conda env create -f environment.yml

Then activate the environment: conda activate cil2_demos

There are 3 directories for 3 different case studies:

• CaseStudy_ColourProcessing (Section 3) :

  1. Color Denoising
  2. Color Inpainting
     
     

• CaseStudy_DynamicTomography (Section 4) :

  1. 01_LoadData_CreateSparseData
  2. 02_FBP_reconstructions
  3. 03_TikhonovReconstructions
  4. 04_TVReconstructions
  5. 05_dTVReconstructions
  6. 06_ShowFigures
     
     

• CaseStudy_HyperspectralTomography (Section 5) :

  1. 01_LoadRawDataAndCrop
  2. 02_PreProcessRingRemover
  3. 03_SIRT_reconstructions
  4. 04_SPDHG_SpatioSpectralTV
  5. 05_SPDHG_3D_spectral_TV
  6. 06_PDHG_SpatioSpectralTV
  7. 07_ShowFigures
     
     

Reference

Papoutsellis E et al. 2021 Core imaging library part II: multichannel reconstruction for dynamic and spectral tomography. Phil. Trans. R. Soc. A 20200193. https://doi.org/10.1098/rsta.2020.0193, preprint