![Hier klicken, um den Treffer aus der Auswahl zu entfernen](images/unchecked.gif) |
Titel |
Parameter Estimation of Fossil Oysters from High Resolution 3D Point Cloud and Image Data |
VerfasserIn |
Ana Djuricic, Mathias Harzhauser , Peter Dorninger, Clemens Nothegger, Oleg Mandic, Balázs Székely, Gábor Molnár, Norbert Pfeifer |
Konferenz |
EGU General Assembly 2014
|
Medientyp |
Artikel
|
Sprache |
Englisch
|
Digitales Dokument |
PDF |
Erschienen |
In: GRA - Volume 16 (2014) |
Datensatznummer |
250100151
|
Publikation (Nr.) |
EGU/EGU2014-16040.pdf |
|
|
Schlagwörter |
Fossilien, Bivalvia, Ostreidae, Fossilfundstelle, Austernriff, Modellierung, 3D-Modell, Visualisierung, Korneuburger Becken |
Geograf. Schlagwort |
Österreich, Niederösterreich, Korneuburg (Bezirk), Stetten |
Blattnummer |
41 [Deutsch Wagram] |
Blattnummer (UTM) |
5320 [Wien] |
|
|
|
Zusammenfassung |
A unique fossil oyster reef was excavated at Stetten in Lower Austria, which is also the
highlight of the geo-edutainment park “Fossilienwelt Weinviertel”. It provides the rare
opportunity to study the Early Miocene flora and fauna of the Central Paratethys Sea. The site
presents the world’s largest fossil oyster biostrome formed about 16.5 million years ago in a
tropical estuary of the Korneuburg Basin. About 15,000 up to 80-cm-long shells of
Crassostrea gryphoides cover a 400 m2 large area. Our project “Smart-Geology for the
World’s largest fossil oyster reef” combines methods of photogrammetry, geology and
paleontology to document, evaluate and quantify the shell bed. This interdisciplinary
approach will be applied to test hypotheses on the genesis of the taphocenosis (e.g.: tsunami
versus major storm) and to reconstruct pre- and post-event processes. Hence, we are
focusing on using visualization technologies from photogrammetry in geology
and paleontology in order to develop new methods for automatic and objective
evaluation of 3D point clouds. These will be studied on the basis of a very dense
surface reconstruction of the oyster reef. “Smart Geology”, as extension of the
classic discipline, exploits massive data, automatic interpretation, and visualization.
Photogrammetry provides the tools for surface acquisition and objective, automated
interpretation.
We also want to stress the economic aspect of using automatic shape detection in
paleontology, which saves manpower and increases efficiency during the monitoring and
evaluation process. Currently, there are many well known algorithms for 3D shape detection
of certain objects. We are using dense 3D laser scanning data from an instrument utilizing the
phase shift measuring principle, which provides accurate geometrical basis < 3 mm.
However, the situation is difficult in this multiple object scenario where more than 15,000
complete or fragmentary parts of an object with random orientation are found. The goal is to
investigate if the application of state-of-the-art 3D digitizing, data processing, and
visualization technologies support the interpretation of this paleontological site. The obtained
3D data (approx. 1 billion points at the respective area) is analyzed with respect to their 3D
structure in order to derive geometrical information. The aim of this contribution is to
segment the 3D point cloud of laser scanning data into meaningful regions representing
particular objects.
Geometric parameters (curvature, tangent plane orientation, local minimum and maximum,
etc.) are derived for every 3D point of the point cloud. A set of features is computed in each
point using different kernel sizes to define neighborhoods of different size. This provides
information on convexity (outer surface), concavity (inner surface) and locally flat areas,
which shall be further utilized in fitting model of Crassostrea-shells. In addition, digitizing is
performed manually in order to obtain a representative set of reference data for the evaluation
of the obtained results. For evaluating these results the reference data (length and orientation
of specimen) is then compared to the automatically derived segments of the point
cloud.
The study is supported by the Austrian Science Fund (FWF P 25883-N29). |
|
|
|
|
|