 |
| Titel |
The Pore3D software library applied to the quantitative morphological and textural analysis of three-dimensional images in geosciences |
| VerfasserIn |
Lucia Mancini, Francesco Brun, Diego Dreossi, Georgios Kourousias, Margherita Polacci, Marco Voltolini, Giuliana Tromba |
| Konferenz |
EGU General Assembly 2011
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 13 (2011) |
| Datensatznummer |
250046513
|
|
|
|
|
|
| Zusammenfassung |
| Imaging techniques play an important role in several research fields of the geosciences. As an
example, the well-known scanning electron microscopy (SEM) technique is a widely
adopted tool for the investigation of the texture of volcanic rock samples. X-ray based
techniques are also of particular interest and, for instance, X-ray microradiography has
proved to be useful for the monitoring of the heavy-metal accumulation in vegetal
tissues.
In recent years great interest has been posed in tomographic techniques like X-ray
computed microtomography (μ-CT) and the micro-scale magnetic resonance imaging
(μ-MRI) as they produce three-dimensional (3D) images of the internal structure of objects.
In fact, investigations performed directly in the 3D domain overcome the limitations of
stereological methods usually applied on SEM-based analysis. Moreover, μ-CT and
μ-MRI techniques enable to get 3D images of the internal core of a sample in a non
destructive way and with a spatial resolution in the micro-meter scale more suitable
for further quantitative analysis. An intriguing challenge lies on the extraction of
quantitative measures and indices directly from these kinds of images. Porosity and
specific surface area as well as anisotropy, connectivity and tortuosity are interesting
descriptors of a 3D model. However, accurate image processing and analysis methods
for an effective assessment of these parameters are still an open issue in several
applications.
The present work illustrates the Pore3D software library developed by the SYRMEP
research group of the Elettra Synchrotron Light Laboratory in Trieste (Italy). Although any
kind of 2D image and 3D dataset is a valid input, Pore3D has been conceived for the handling
of X-ray μ-CT images. In fact, the particular case of high-resolution μ-CT images requires ad
hoc software tools able to manage large 3D dataset in an easy way. Several commercial
software and public available libraries as well as research codes have been developed in
recent years for 2D and 3D image analysis. Within this framework the Pore3D project
merges many of the features implemented in existing software tools, customizing in
some cases their characteristics or adding new tools, on the basis of the specific
know-how acquired inside the SYRMEP collaboration. Moreover, a custom developed
software assures a complete control of the algorithm implementation and permits
different strategies of analysis as a function of the specific scientific application. The
current version of Pore3D consists in a library of state-of-the-art functions and
procedures for filtering, segmentation and quantitative analysis of images. It is
also under study the deployment of Pore3D according to the software-as-a-service
paradigm.
In this work we present several applications of the Pore3D software to the extraction of
quantitative information from μ-CT images of geomaterials. In particular, some results
obtained in the characterization of texture and 3D shape orientation distribution of crystals
and vesicles in volcanic products will be illustrated. |
|
|
|
|
|
|