Classification of Wetland Vegetation with Multi-temporal Dual Frequency SAR data

Abstract

Wetlands is one most valuable ecosystems on the earth. Also today the importance of wetland is pointed out as a major emission source of methane which is one of the green house gases.
The development of remote sensing method to monitor wetlands vegetation distribution is urgent necessity. The purpose of this study is to investigate remote sensing methodology for monitoring wetland vegetation distribution by utilizing Synthetic Aperture Radar data. Vegetation classification was performed with multi-temporal JERS-1/SAR (L band) and ERS-1/AMI (C band) images. As initial step for classification, the speckle noise reduction was performed by MAP filter. As the result of supervised classification by Maximum likelihood method, an accurate wetland vegetation classification map was produced.