Multimodal Visualization and Analysis of Spectral and XCT Data

Bernhard Fröhler, Artem Amirkhanov, Johann Kastner, Meister Eduard Gröller, Christoph Heinzl
Multimodal Visualization and Analysis of Spectral and XCT Data
Computer Graphic Forum, 33(3):91-100, April 2015.

Information

Abstract

An increasing number of industrial applications demand a comprehensive analysis of both structural and chemical composition. Typically, non-destructive testing techniques focus on either structural or chemical characterization but do not deliver both. 3D X-Ray Computed Tomography (XCT) scans are well-suited for determining the internal and external structure of an object at high resolution. The attenuation value it delivers can however be the same or very similar for different materials. For a detailed chemical analysis XCT is therefore combined with spectral characterization techniques such as K-Edge Absorptiometry or X-ray Fluorescence Spectroscopy. In this paper, we are extending a previously introduced framework for visualization and analysis of specimens scanned with these two modalities in multiple ways: For better understanding the dependencies between the spectral energy levels, we propose Spectral Similarity Maps. Spectral Functional Boxplots visualize the statistical distribution of the spectral data. The Spectrum Explor-er improves the analysis of specimens of unknown composition. We demonstrate the usefulness of our techniques on several use cases.

Additional Files and Images

No additional files or images.

Weblinks

BibTeX

@article{Froehler_Berhnard_2015_ESM,
  title =      "Multimodal Visualization and Analysis of Spectral and XCT
               Data",
  author =     "Bernhard Fr{"o}hler and Artem Amirkhanov and Johann Kastner
               and Meister Eduard Gr{"o}ller and Christoph Heinzl",
  year =       "2015",
  abstract =   "An increasing number of industrial applications demand a
               comprehensive analysis of both structural and chemical
               composition. Typically, non-destructive testing techniques
               focus on either structural or chemical characterization but
               do not deliver both. 3D X-Ray Computed Tomography (XCT)
               scans are well-suited for determining the internal and
               external structure of an object at high resolution. The
               attenuation value it delivers can however be the same or
               very similar for different materials. For a detailed
               chemical analysis XCT is therefore combined with spectral
               characterization techniques such as K-Edge Absorptiometry or
               X-ray Fluorescence Spectroscopy. In this paper, we are
               extending a previously introduced framework for
               visualization and analysis of specimens scanned with these
               two modalities in multiple ways: For better understanding
               the dependencies between the spectral energy levels, we
               propose Spectral Similarity Maps. Spectral Functional
               Boxplots visualize the statistical distribution of the
               spectral data. The Spectrum Explor-er improves the analysis
               of specimens of unknown composition. We demonstrate the
               usefulness of our techniques on several use cases.",
  month =      apr,
  issn =       "2411-5428",
  journal =    "Computer Graphic Forum",
  note =       "appeared in June 2014",
  number =     "3",
  volume =     "33",
  pages =      "91--100",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2015/Froehler_Berhnard_2015_ESM/",
}