2
International Symposium on Digital Industrial Radiology and Computed Tomography – DIR2019 pyXIT: Python GUI Software for CT Imaging, Learning and Experimental Algorithms; Example: Multi Material Phase Retrieval Maximilian ULLHERR  1 , Simon ZABLER  2 , Jonas GRÄTZ  1 , Randolf HANKE  3 1  Universität Würzburg, Würzburg, Germany 2  Fraunhofer EZRT, Würzburg, Germany 3  Fraunhofer EZRT, Fürth, Germany Contact e-mail: [email protected] Abstract The Python X-Ray Imaging Tool (pyXIT) is a software for CT reconstruction and image processing. It is open and free software (BSD license) [1]. pyXIT is easily extendable with plugins written in Python and includes a basic CT reconstruction algorithm (cone FBP). The processing speed is comparable to commercial software and it is designed to process many (> 20) and large data sets (4K) comfortably. The software runs in a Python environment (≥ v3.6) with the usual modules for scientific computing installed and has no other dependencies. pyXIT is not meant to replace commercial software, but it is rather an environment for learning CT and using own or experimental code. It is intended to be used by an experienced user. The original use case is for reconstructing high-resolution CTs at the chair for X-Ray microscopy of the University Würzburg. Some features developed for this use case are: phase retrieval, field of view extension, drift correction (from rewind images), lens distortion correction. Plugins for image processing can easily be integrated into pyXIT and the preview allows to test/debug or parameterize an algorithm. One example for such a plugin is our implementation of a multi material phase retrieval algorithm [2]. It is a volume filter which applies phase retrieval for inline phase contrast in multiple material samples. As a framework, pyXIT provides many features which simplify plugin development or can be used for independent projects. The most important are: GPU computing with OpenCl (own API layer), multiprocessing, Fourier filters, file access, preview. [1] physik.uni-wuerzburg.de/lrm/forschung/software/ [2] Ullherr, M, and Zabler, S.. "Correcting multi material artifacts from single material phase retrieved holo-tomograms with a simple 3D Fourier method." Optics express (2015), DOI: 10.1364/OE.23.032718 1 License: This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA. More info about this article: http://www.ndt.net/?id=24774

pyXIT: Python GUI Software for CT Imaging, Learning and ...The Python X-Ray Imaging Tool (pyXIT) is a software for CT reconstruction and image processing. It is open and free software

  • Upload
    others

  • View
    20

  • Download
    0

Embed Size (px)

Citation preview

International Symposium on Digital Industrial Radiology and Computed Tomography – DIR2019

pyXIT: Python GUI Software for CT

Imaging, Learning and Experimental

Algorithms; Example: Multi Material Phase

Retrieval

Maximilian ULLHERR 1, Simon ZABLER 2, Jonas GRÄTZ 1, Randolf HANKE 3

1 Universität Würzburg, Würzburg, Germany2 Fraunhofer EZRT, Würzburg, Germany

3 Fraunhofer EZRT, Fürth, Germany

Contact e-mail: [email protected]

Abstract

The Python X-Ray Imaging Tool (pyXIT) is a software for CT reconstruction and imageprocessing. It is open and free software (BSD license) [1]. pyXIT is easily extendable withplugins written in Python and includes a basic CT reconstruction algorithm (cone FBP).The processing speed is comparable to commercial software and it is designed to processmany (> 20) and large data sets (4K) comfortably. The software runs in a Pythonenvironment (≥ v3.6) with the usual modules for scientific computing installed and has noother dependencies.

pyXIT is not meant to replace commercial software, but it is rather an environmentfor learning CT and using own or experimental code. It is intended to be used by anexperienced user. The original use case is for reconstructing high-resolution CTs at thechair for X-Ray microscopy of the University Würzburg. Some features developed for thisuse case are: phase retrieval, field of view extension, drift correction (from rewind images),lens distortion correction.

Plugins for image processing can easily be integrated into pyXIT and the previewallows to test/debug or parameterize an algorithm. One example for such a plugin is ourimplementation of a multi material phase retrieval algorithm [2]. It is a volume filter whichapplies phase retrieval for inline phase contrast in multiple material samples.

As a framework, pyXIT provides many features which simplify plugin developmentor can be used for independent projects. The most important are: GPU computing withOpenCl (own API layer), multiprocessing, Fourier filters, file access, preview.

[1] physik.uni-wuerzburg.de/lrm/forschung/software/[2] Ullherr, M, and Zabler, S.. "Correcting multi material artifacts from single

material phase retrieved holo-tomograms with a simple 3D Fourier method." Optics express(2015), DOI: 10.1364/OE.23.032718

1License: This work is licensed under the Creative Commons Attribution 4.0 International License. To view a copy of this license, visithttp://creativecommons.org/licenses/by/4.0/ or send a letter to Creative Commons, PO Box 1866, Mountain View, CA 94042, USA.

Mor

e in

fo a

bout

this

art

icle

: ht

tp://

ww

w.n

dt.n

et/?

id=

2477

4

2