Comparison of IEM and SPM Model for Oil Spill Detection Using Inversion Technique and Radar Data

The purpose of this work is to verifying L band wave behavior for oil spill detection over oceans from satellite remote sensing sensors. Moreover, evaluation of the optimum theoretical backscattering model for this application is considered. This objective was realized by performing an inversion tec...

Full description

Saved in:
Bibliographic Details
Published in:Journal of the Indian Society of Remote Sensing 2013-06, Vol.41 (2), p.425-431
Main Authors: Shahsavarhaghighi, Safoora, Sahebi, Mohmod Reza, Valdanzoej, Mohammad Javad, Haddadi, G. Ataollah
Format: Article
Language:English
Subjects:
algorithms
data collection
dielectric properties
Earth and Environmental Science
Earth Sciences
equations
Marine
neural networks
oceans
oil spills
oils
radar
remote sensing
Remote Sensing/Photogrammetry
Research Article
roughness
surface roughness
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The purpose of this work is to verifying L band wave behavior for oil spill detection over oceans from satellite remote sensing sensors. Moreover, evaluation of the optimum theoretical backscattering model for this application is considered. This objective was realized by performing an inversion technique which uses neural networks and backscattering model to estimate ocean surface parameters like roughness and dielectric constant. The normalized measure of the radar signal that backscattered to the antenna (sigma zero) includes specific characteristics of the target such as roughness and dielectric constant. Although, in radar imagery, surface roughness is the dominant factor in determining the amplitude of the return signal, this study, most often focus on the dielectric properties because there is a significant difference between dielectric constant of water and oil and therefore they can be distinguished easily by estimating dielectric constant parameter. The neural networks were first trained with a simulated data set generated from the integral equation model (IEM) and second with metadata from small perturbation model (SPM). It used a fast learning algorithm for training a multilayer feed forward neural network. A theoretical database was used for the learning stage. The proposed approach was applied with ALOS-PALSAR images from the Philippine Sea and output are presented in binary images.
ISSN:0255-660X
0974-3006
DOI:10.1007/s12524-012-0217-4