Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal LAN0710

TitleMultisource SVM Classification of Moorea Landcover using TerraSARX, Quickbird and DEM Data
Investigator Stoll, Benoit - French Polynesia University, GePaSud Laboratory
Team Member
Dr. Stoll, Benoit - GePaSud Lab. French Polynesia University, Remote Sensing and Image Processing
Dr. Chabrier, Sebastien - GePaSud Laboratory, French Polynesia University, Remote Sensing and Image Processing
PhD Student Pouteau, Robin - GePaSud Laboratory, French Polynesia University, Remote Sensing and Image Processing
Dr. Meyer, Jean-Yves - French Polynesia Government, Ministere de lEducation, de lEnseignement Superieur, Research Department
SummaryThis project is part of the ambitious international program Moorea Biocode Project http://www.mooreabiocode.org/ , it aims to produce high definition classification maps of the vegetation covers of Moorea island in order to optimize the ground truth mission that will be conducted by the biologists thus improving the ground truth paths and sampling. The new SVM algorithms will be used to classify high definition multi-sources data: optical (Quickbird) radar (TerraSARX) and DEM derived topographic indexes. We aim to classify as accurately as possible the different vegetation covers and land covers using the complementary properties of the radar data, optical data and topographic indexes which proved to improve natural landscape covers classification.

Back to list of proposals

DLR 2004-2016