Segmentation of X-ray CT data of porous materials: A review of global and locally adaptive algorithms

Markus Tuller, Ramaprasad Kulkarni, Wolfgang Fink

Research output: Chapter in Book/Report/Conference proceedingChapter

23 Scopus citations


Recent computational and technological advances in X-ray computed tomography (CT) provide exciting new means for nondestructive, three-dimensional imaging of soil-water-root processes with projects ranging from visualization and characterization of biofilms to quantification of effects of root-induced compaction on rhizosphere hydraulic properties. In contrast to these breath-taking technological improvements, the development and evaluation of appropriate segmentation methods for transformation of three-dimensional grayscale X-ray CT data into a discrete form that allows accurate separation of solid, liquid, and vapor phases for quantitative description of the soil-water-root continuum and provides the basis for modeling of static and dynamic system processes appears to lag behind. This chapter reviews the state-of-the-art and recent developments in image segmentation, covering global and locally adaptive two-phase as well as more complex multiphase algorithms. The need for true three-dimensional segmentation is demonstrated and potential techniques for pre-segmentation enhancement of X-ray CT data are discussed.

Original languageEnglish (US)
Title of host publicationSoil- Water- Root Processes
Subtitle of host publicationAdvances in Tomography and Imaging
Number of pages26
ISBN (Electronic)9780891189596
ISBN (Print)9780891189589
StatePublished - Jan 1 2015


  • Global adaptive algorithm
  • Locally adaptive algorithm
  • Porous material
  • Segmentation algorithm
  • Soil science research
  • Three-dimensional grayscale X-ray CT data
  • X-ray computed tomography

ASJC Scopus subject areas

  • General Engineering
  • General Agricultural and Biological Sciences


Dive into the research topics of 'Segmentation of X-ray CT data of porous materials: A review of global and locally adaptive algorithms'. Together they form a unique fingerprint.

Cite this