LAUSR

ABSTRACT Fully polarimetric synthetic aperture radar (SAR) has been widely used in applications, such as ground parameter inversion and classification. However, to accurately decipher the information contained in the polarimetric characteristics of an illuminated scene, the cross talks in cross-polarimetric channels and imbalances among co-pol channels must be removed from the measured polarimetric responses. Therefore, it is important to improve the polarimetric SAR calibration technique to reduce the distortions between various channels and calibrate the polarimetric data. In this article, we propose an efficient polarimetric SAR calibration method that uses known calibration point targets to remove the uncertainties among observed polarized responses. We first construct a common scattering model of natural targets organized by matrix products. Next, we reorder the square matrices in vector form. Finally, we construct a set of analytical equations based on the relations between theoretical and measured scattering matrices and derive the full elements of the distortion matrix by solving these equations. This algorithm determines the calibration matrix from corner reflectors with known scattering matrices, and it is efficient and easy to understand. Finally, we validate the algorithm with airborne C-band polarimetric SAR data.