Details
| Originalsprache | Englisch |
|---|---|
| Seiten (von - bis) | 421-434 |
| Seitenumfang | 14 |
| Fachzeitschrift | Photogrammetrie, Fernerkundung, Geoinformation |
| Jahrgang | 2011 |
| Ausgabenummer | 6 |
| Publikationsstatus | Veröffentlicht - 1 Dez. 2011 |
Abstract
In this paper we present an approach for automatic coastline detection from images based on snakes (parametric active contours) and apply it to Landsat images from Antarctica. Snakes require the definition of an energy functional that reflects the underlying coastline model. For Antarctica the coastline appearance in the images is heterogeneous. Therefore, it is not possible to use a single model only. After inspecting the images to be used we formulate three different transition models that match a large part of the Antarctic coastline: (a) from ice shelf to water, (b) from ice shelf to sea ice and (c) from rocky terrain to water. For each of the three models the energy terms are optimised based on the radiometric properties of the adjacent regions as well as the curvature and the potential change-rate of the coastline itself. A supervised classification for the three classes ice, water and rocky terrain controls the whole process by selecting the most applicable model for a given image region along the coastline. We present results for the extraction of approximately 12%of theAntarctic coastline from an up-to-date Landsat mosaic.
ASJC Scopus Sachgebiete
- Sozialwissenschaften (insg.)
- Geografie, Planung und Entwicklung
- Physik und Astronomie (insg.)
- Instrumentierung
- Erdkunde und Planetologie (insg.)
- Erdkunde und Planetologie (sonstige)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
in: Photogrammetrie, Fernerkundung, Geoinformation, Jahrgang 2011, Nr. 6, 01.12.2011, S. 421-434.
Publikation: Beitrag in Fachzeitschrift › Artikel › Forschung › Peer-Review
}
TY - JOUR
T1 - Antarctic Coastline Detection using Snakes
AU - Klinger, Tobias
AU - Ziems, Marcel
AU - Heipke, Christian
AU - Schenke, Hans Werner
AU - Ott, Norbert
PY - 2011/12/1
Y1 - 2011/12/1
N2 - In this paper we present an approach for automatic coastline detection from images based on snakes (parametric active contours) and apply it to Landsat images from Antarctica. Snakes require the definition of an energy functional that reflects the underlying coastline model. For Antarctica the coastline appearance in the images is heterogeneous. Therefore, it is not possible to use a single model only. After inspecting the images to be used we formulate three different transition models that match a large part of the Antarctic coastline: (a) from ice shelf to water, (b) from ice shelf to sea ice and (c) from rocky terrain to water. For each of the three models the energy terms are optimised based on the radiometric properties of the adjacent regions as well as the curvature and the potential change-rate of the coastline itself. A supervised classification for the three classes ice, water and rocky terrain controls the whole process by selecting the most applicable model for a given image region along the coastline. We present results for the extraction of approximately 12%of theAntarctic coastline from an up-to-date Landsat mosaic.
AB - In this paper we present an approach for automatic coastline detection from images based on snakes (parametric active contours) and apply it to Landsat images from Antarctica. Snakes require the definition of an energy functional that reflects the underlying coastline model. For Antarctica the coastline appearance in the images is heterogeneous. Therefore, it is not possible to use a single model only. After inspecting the images to be used we formulate three different transition models that match a large part of the Antarctic coastline: (a) from ice shelf to water, (b) from ice shelf to sea ice and (c) from rocky terrain to water. For each of the three models the energy terms are optimised based on the radiometric properties of the adjacent regions as well as the curvature and the potential change-rate of the coastline itself. A supervised classification for the three classes ice, water and rocky terrain controls the whole process by selecting the most applicable model for a given image region along the coastline. We present results for the extraction of approximately 12%of theAntarctic coastline from an up-to-date Landsat mosaic.
KW - Antarctica
KW - Automation
KW - Classification
KW - Coastline
KW - Model selection
KW - Snakes (active contours)
UR - http://www.scopus.com/inward/record.url?scp=84855233145&partnerID=8YFLogxK
U2 - 10.1127/1432-8364/2011/0095
DO - 10.1127/1432-8364/2011/0095
M3 - Article
AN - SCOPUS:84855233145
VL - 2011
SP - 421
EP - 434
JO - Photogrammetrie, Fernerkundung, Geoinformation
JF - Photogrammetrie, Fernerkundung, Geoinformation
SN - 1432-8364
IS - 6
ER -