Title :
Scattering mechanism identification based on the rotation and eccentric angles of polarimetric SAR data
Author_Institution :
Dept. of Space & Earth Inf. Technol., Kyushu Tokai Univ.
Abstract :
In this paper we introduce two novel parameters for polarimetric SAR (POLSAR) images. For this purpose, a 2times2 matrix is defined to take into account the independent elements of the covariance matrix. To understand the geometry of this new matrix, an equation of ellipse is defined. As a result, the geometry of the ellipse is completely tied up with the eigenvalues and eigenvectors. Two novel parameters, the rotation and eccentric angles of the ellipse which can directly be derived from the eigenvalues and eigenvectors, are presented. The results indicate that different scattering mechanisms can be represented with the rotation and eccentric angles. Results obtained from the proposed parameters, when compared with the results derived from existing features, were in agreement in terms of the scattering mechanisms, as well as terrain cover types present in the test site
Keywords :
backscatter; covariance matrices; eigenvalues and eigenfunctions; geophysical signal processing; geophysical techniques; radar imaging; radar polarimetry; remote sensing by radar; synthetic aperture radar; POLSAR images; covariance matrix; eigenvalues; eigenvectors; ellipse equation; elliptical geometry; polarimetric SAR data; polarimetric SAR images; polarimetric features; radar polarimetry; scattering mechanism identification; Covariance matrix; Eigenvalues and eigenfunctions; Entropy; Equations; Geometry; Polarization; Radar scattering; Scattering parameters; Space technology; Testing;
Conference_Titel :
Geoscience and Remote Sensing Symposium, 2004. IGARSS '04. Proceedings. 2004 IEEE International
Conference_Location :
Anchorage, AK
Print_ISBN :
0-7803-8742-2
DOI :
10.1109/IGARSS.2004.1370342