 |
| Titel |
Image analysis of dye stained patterns in soils |
| VerfasserIn |
Christina Bogner, Baltasar Trancón y Widemann, Holger Lange |
| Konferenz |
EGU General Assembly 2013
|
| Medientyp |
Artikel
|
| Sprache |
Englisch
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 15 (2013) |
| Datensatznummer |
250077397
|
|
|
|
|
|
| Zusammenfassung |
| Quality of surface water and groundwater is directly affected by flow processes in the
unsaturated zone. In general, it is difficult to measure or model water flow. Indeed,
parametrization of hydrological models is problematic and often no unique solution exists. To
visualise flow patterns in soils directly dye tracer studies can be done. These experiments
provide images of stained soil profiles and their evaluation demands knowledge in hydrology
as well as in image analysis and statistics.
First, these photographs are converted to binary images classifying the pixels in dye
stained and non-stained ones. Then, some feature extraction is necessary to discern relevant
hydrological information. In our study we propose to use several index functions to
extract different (ideally complementary) features. We associate each image row with
a feature vector (i.e. a certain number of image function values) and use these
features to cluster the image rows to identify similar image areas. Because images of
stained profiles might have different reasonable clusterings, we calculate multiple
consensus clusterings. An expert can explore these different solutions and base
his/her interpretation of predominant flow mechanisms on quantitative (objective)
criteria.
The complete workflow from reading-in binary images to final clusterings has
been implemented in the free R system, a language and environment for statistical
computing. The calculation of image indices is part of our own package Indigo,
manipulation of binary images, clustering and visualization of results are done
using either build-in facilities in R, additional R packages or the LATEX system. |
|
|
|
|
|
|