Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → Israel → SAR backscatter dependence on various land covers in semi-arid terrain

Ben Gurion University of the Negev (2014)

SAR backscatter dependence on various land covers in semi-arid terrain

Hetz Marina

Titre : SAR backscatter dependence on various land covers in semi-arid terrain

Auteur : Hetz Marina

Université de soutenance : Ben Gurion University of the Negev

Grade : Doctor of Philosophy (PhD) 2014

Résumé partiel
Assessing and monitoring the state of the Earth is a key requirement for global change research. Sustainable management of any ecosystem requires, among other information, a thorough understanding of vegetation species distribution. Classifying and mapping vegetation is an important technical task for managing natural resources as vegetation provides a basis for all living beings and plays an essential role in affecting global climate change, such as influencing terrestrial CO2. Vegetation mapping also presents valuable information for understanding natural and man-made environments through quantifying vegetation cover from local to global scales at a given time point or over a continuous period. Nowadays, remote sensing is a major source for spatial information regarding the Earth’s surface cover and vegetation ; it can provide information that is not currently part of an existing forest inventory. The technology of remote sensing offers a practical and economical means to study vegetation cover changes, especially over large areas. Radar remote sensing refers to the long wavelengths of the electromagnetic spectrum [ 1-130cm]. At these wavelengths, the atmosphere is completely transparent, making the use of optical sensors impossible, and radar becomes a useful tool. Irrespective to wavelength, radar signals can be transmitted and/or received in different modes of polarization. Four possible combinations of signal transmission and reception exist : HH, HV, VH, and VV. As quad-pol imagery is not always available and decomposition methods for terrain classification are rarely available, an alternate way to classify radar images is required. Dual-pol systems provide images in both horizontal and vertical polarizations (HH and VV). Most systems do not have these capabilities, and, if they do, one of the polarizations may fail. The main goal of the current research was to examine the ability to classify different vegetation types based on a single-polarized radar image. By definition, a single-band image cannot provide a spectral-based segmentation and classification. Thus, because most available SAR images are single-band, a straightforward classification cannot be applied on such data. In this study, we have demonstrated how statistical parameters related to the clutter of different vegetation types can be used to segment and classify vegetation in single-band SAR images.

Mots clés : Backscattering – Experiments — Remote sensing — Arid regions — Synthetic aperture radar — Image quality

Présentation

Version intégrale

Page publiée le 9 novembre 2016, mise à jour le 12 septembre 2017