For Full-Text PDF, please login, if you are a member of IEICE,|
or go to Pay Per View on menu list, if you are a nonmember of IEICE.
Use of Multi-Polarimetric Enhanced Images in SIR-C/X-SAR Land-Cover Classification
Takeshi NAGAI Yoshio YAMAGUCHI Hiroyoshi YAMADA
IEICE TRANSACTIONS on Communications
Publication Date: 1997/11/25
Print ISSN: 0916-8516
Type of Manuscript: PAPER
Category: Measurement and Metrology
SAR image analysis, radar polarimetry, classification, wavelet transform, remote sensing,
Full Text: PDF(1.1MB)
>>Buy this Article
This paper presents a method for land cover classification using the SIR-C/X-SAR imagery based on the maximum likelihood method and the polarimetric filtering. The main feature is to use polarimetric enhanced image information in the pre-processing stage for the classification of SAR imagery. First, polarimetric filtered images are created where a specific target is enhanced versus another, then the image data are incorporated into the feature vector which is essential for the maximum likelihood classification. Specific target classes within the SAR image are categorized according to the maximum likelihood method using the wavelet transform. Addition of polarimetric enhanced image in the preprocessing stage contributes to the increase of classification accuracy. It is shown that the use of polarimetric enhanced images serves efficient classifications of land cover.