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

Accueil du site → Master → Canada → Crop discrimination and monitoring with multi-frequency and multi-polarized SAR data : a dryland farming example

Carleton University (1990)

Crop discrimination and monitoring with multi-frequency and multi-polarized SAR data : a dryland farming example

Hutton, Christine A.,

Titre : Crop discrimination and monitoring with multi-frequency and multi-polarized SAR data : a dryland farming example

Auteur : Hutton, Christine A

Université de soutenance : Carleton University

Grade : Master of Arts 1990

Résumé
The objective of this paper is to introduce a variation on the standard classification techniques used in the analysis of agricultural land use with Synthetic Aperture Radar (SAR). Following an overview of background leading to SAR research and the basic SAR fundamentals, several issues were considered. A Maximum Likelihood Pixel Classifier, a filter of the Pixel Classifier results and a Field Classifier were the techniques used. The methodologies used were designed to investigate the effects of frequency and polarization on : (1) the capability to discriminate crop type with C-, X- and L-band SAR, (2) the classification accuracies possibilities with single band SAR and numerous combinations of frequency and polarization and (3) the effects of crop row orientation on the potential to classify the crop type. The SAR data used were acquired over a dryland farming region near the city of Swift Current Saskatchewan in 1984. Alternating strips of wheat and summerfallow comprise the major ground cover types. The results showed that only C-VH data were an adequate single band data source and that greater accuracies were found as the number of band and polarization combinations was increased.

Présentation

Version intégrale

Page publiée le 21 mars 2010, mise à jour le 9 février 2018