Calibration of polarimetric SAR data subject to Faraday rotation using a Genetic algorithm

Recently, some spaceborne polarimetric synthetic aperture radars, such as TerraSAR X (Xband), RADARSAR-2 (C-band) and PALSAR (L-band), are available to measure a scattering matrix of terrain and polarimetric data analysis techniques are being developed for terrain classification, forest biomass and soil moisture estimations, etc. Thus polarimetric calibration is necessary to obtain the accurate results form the polarimetric SAR data [1]-[3]. In the case of spaceborne SAR, Faraday rotation affects a measurement of the polarimetric SAR data and its polarimetric calibration becomes complicated procedure. The calibration parameters are channel imbalance, cross-talk and Faraday rotation angle. Freeman proposed the polarimetric calibration method to estimate channel imbalance and Faraday rotation angle only under the condition that cross-talk is negligible and terrain has the reflection symmetry property [4]. In this paper, we propose a polarimetric calibration method to estimate all calibration parameters using a genetic algorithm [5] and polarimetric SAR data does not need to satisfy the reflection symmetry. In section 2, polarimetric calibration model affected by Faraday rotation is addressed. Next, polarimetric calibration method using the genetic algorithm is explained. Finally, we examine that the proposed method is applied to the actual polarimetric SAR data.