Skip to content

Code associated with the paper 'Directional Sinogram Inpainting for Limited Angle Tomography' currently on arXiv.

Notifications You must be signed in to change notification settings

ChristophBrune/2018_Directional_Inpainting_for_Limited_Angle

 
 

Repository files navigation

Directional Sinogram Inpainting for Limited Angle Tomography
Contact: Robert Tovey ([email protected])
Other Authors: Martin Benning, Carola-Bibiane Schoenlieb, Sean M. Collins,
Rowan K. Leary, Paul A. Midgley, Christoph Brune, Marinus J. Lagerwerf.

============================================================

This repository contains the code required to generate reconstructions 
for the figures in the relevant paper (see paper.pdf). To run with alternative
phantoms/settings in Exp4/5 simply uncomment/alter parameters in the first 
section of the script file.

This repository comes with the AIR Tools package (Per Christian Hansen, 
https://www.sciencedirect.com/science/article/pii/S0377042711005188) for 
convenience which provides the SIRT algorithm implementation. 

The copyright for all work within this repository (minus the AIR Tools package)
is the Creative Commons Attribution 4.0 International (CC BY 4.0) license.

Requirements are MATLAB (tested on 2016b) with CVX addon (http://cvxr.com/cvx/)
and the astra toolbox (https://www.astra-toolbox.com/). We recommend using the 
Mosek solver for CVX, free academic licenses are available 
(http://cvxr.com/cvx/doc/mosek.html). Any external libraries can be added to 
the path using the startup script.

Included scripts are:
startup: Adds directories to path. Must be run before other scripts.
Exp1: Generates Figure 7
Exp2: Generates both plots of Figure 4
Exp3: Generates Figure 3
Exp4: Generates Figures 10 and 11. Figure 5 is a subplot of Figure 11
Exp5: Generates Figures 13 and 14

About

Code associated with the paper 'Directional Sinogram Inpainting for Limited Angle Tomography' currently on arXiv.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • MATLAB 100.0%