Complémentarité des images optiques et radars pour la connaissance des littoraux

Abstract : A large share of world population lives on sea shore rendering necessary to better study and characterize these areas. Remote sensing from space offers a large potential, especially when combining various sources of images. The present work deals with the mapping of properties of the sea shore. Under certain conditions (storms, tsunami...), the coastline may evolve dramatically. Imaging radars have capabilities to observe such situations. A method is proposed to map the coastline by exploiting image texture. It is assessed by comparison with other methods on a site whose coastline is accurately known, and then on other sites. Optical sensors are also used in this work. Current processing methods extract only one variable relating to littoral. The present work proposes a more global method that extracts several variables. Using a model of the radiative transfer within coastal waters, variables relating to depth, turbidity and sea bed are extracted. The method is assessed against numerical simulations. Then it is applied to actual cases using images acquired by airborne spectrometer and spaceborne radiometer. These examples demonstrate promising results. The present work demonstrates the complementarity of spaceborne sensors for the study of the sea shore. It is shown that accurate and complete maps are best obtained using images acquired by several sensors at different dates.
Complete list of metadatas

https://pastel.archives-ouvertes.fr/pastel-00979423
Contributor : Lucien Wald <>
Submitted on : Wednesday, April 16, 2014 - 7:51:34 AM
Last modification on : Tuesday, March 27, 2018 - 4:06:18 PM
Long-term archiving on : Monday, April 10, 2017 - 2:05:36 PM

Identifiers

  • HAL Id : pastel-00979423, version 1

Citation

Jérôme Bijaoui. Complémentarité des images optiques et radars pour la connaissance des littoraux. Interfaces continentales, environnement. Université de Nice Sophia-Antipolis, 1995. Français. ⟨pastel-00979423⟩

Share

Metrics

Record views

591

Files downloads

457