Belief-propagation for binary tomography reconstruction

This Python code implements belief-propagation iterations for solving the tomography reconstruction problem for binary images with a spatial regularization.

Download

A zipball of the source code is available on https://github.com/eddam/bp-for-tomo/zipball/demo

If you wish to download the whole versioned project, you can clone it with git (http://git-scm.com/):

$ git clone https://github.com/eddam/bp-for-tomo.git

Dependencies

to use this code, you will need the Python language, together with several Python modules

• python >=2.6
• numpy >= 1.4
• scipy
• a C compiler
• optional: matplotlib (to plot the demo results) and scikit-image (for utility functions to preprocess real data).

All these packages are included in the usual Scientific Python distribution, such as Anaconda http://continuum.io/downloads or Enthought Python Distribution https://www.enthought.com/products/epd/.

Install

Go to the bptomo directory

$ cd bptomo

And execute one of the following commands:

If you wish to use the code inside the source directory:

$ python setup.py build_ext --inplace

(this will build the compiled parts of the code, using cython).

Or you can install the package in order to import it from anywhere in the system:

$ python setup.py install --user

Demo

run the script

$ python demo_bp_flavors.py

(or, better, run the script inside the Ipython interpreter)