CCPi-Diamond-Laminography

This repository contains code for processing and understanding the laminography data from the DIAD beamline at Diamond Light Source

Repository structure

The code is arranged in the following folders

Alignment

Notebooks demonstrating the automatic alignment tool designed to align the tilt and centre of rotation offset in laminography data

• laminography_automatic_alignment.ipynb can be used with DIAD laminography data, the user needs to input the filename, an initial guess of tilt and COR, search ranges and tolerance

  file_path = '/mnt/share/ALC_laminography/folder/k11-67208_binned.nxs'
  cor = 0 # pix
  tilt = 35 # deg
  
  optimise_geometry = True
  cor_bounds = (-20, 20) # pixels
  tilt_bounds = (30, 40) # deg
  tilt_tol = 0.01 # deg
  cor_tol = 0.01 # pixels
  

• example_alignment_TEM_grid.ipynb and example_alignment_spheres.ipynb show examples of the scripts use on real datasets

Artefacts

Code demonstrating the kind of artefacts caused by laminography and some methods to combat them

• example_artefacts_grid.ipynb can be run without needing any data to be pre-created
• example_artefacts_cylinder.ipynb requires simulated data generated in the simulations folder

Simulations

Code to create simulated laminography data used in testing the alignment tools and understanding artefacts. Note: code in this folder requires a pre-release version of the gvxr package which is not currently included in the Docker container.

Running the code

To run the code in a Jupyter notebook enabled Docker container on a system with access to a GPU, run the following command:

docker pull ghcr.io/tomographicimaging/ccpi-diamond-laminography:latest
docker run --gpus all -p 8888:8888 ghcr.io/tomographicimaging/ccpi-diamond-laminography:latest