Polarimetric calibration of ultra-wideband SAR imagery

Polarimetric synthetic aperture radar (SAR) imagery is susceptible to degradation because of variations in transmit and receive antenna patterns and hardware limitations on achievable polarimetric isolation. A number of polarimetric calibration procedures have been developed to compensate for these effects. However, most of these techniques are limited to narrowband radar systems. The few that are applicable to ultrawideband (UWB) SAR depend on very high quality antenna measurements or detailed electromagnetic scattering models of test targets. In this thesis we develop and implement a novel blind polarimetric equalization (BPE) procedure suitable for UWB SAR. BPE uses second-order statistics from clutter and targets of opportunity to derive parameters describing the polarimetric response of the radar system. Knowledge of these parameters permits the removal of crosstalk and the correction of imbalances between the polarimetric channels. While not a complete calibration solution, BPE is shown to greatly simplify any subsequent calibration procedures. For UWB applications BPE is implemented as a series of frequency domain operations, and it is this feature that sets the technique apart from previously reported clutter calibration algorithms. The BPE approach was tested and refined on numerical simulations, UWB SAR simulations, and narrowband ISAR measurements. The performance of the technique was then verified on measured LTWB SAR imagery collected by the Army Research Laboratory's BoomSAR system.