4D tomographic reconstruction by time domain decomposition

rectv_gpu on GitHub

How to use

In order to use the rectv environment

  • the enviroment needs to be activated
  • working directory is needed
  • data can be downaloded from tomobank or MAX IV storage
  • a SLURM script may be useful
  • ImageJ can be used for data visualization
# 
# load modules and activate the environment
module purge
module add Anaconda3 CUDA
source activate rectv

Working directory

Use /scratch at MAX IV.

mkdir /scratch/tmp/$USER

Reconstruction script

wget https://raw.githubusercontent.com/math-vrn/rectv_gpu/master/tomopy_rectv.py -O tomopy_rectv.py

SLURM script

wget https://raw.githubusercontent.com/tomograms/tomograms-doc/master/docs/projects/rectv/utils/j_rectv_maxiv.sh -O j_rectv.sh

Start new job:

sbatch j_rectv.sh

Visualizing results (TIFF) with ImageJ

# visualise data with ImageJ
module add  ImageJ

java -jar $EBROOTIMAGEJ/ij.jar -Dij1.plugins.dir=$EBROOTIMAGE/plugins
  • Load an image sequence