Filters for polarimetric calibration of UWB SAR imagery

Polarimetric synthetic aperture radar (SAR) imagery is susceptible to degradation due to non-ideal responses in antennas and other hardware. Image corruption is especially serious in SAR operating over ultra-wide bandwidths (UWB) and imaging over wide angles, as the system response becomes a complicated function of angle and frequency. The usual approach to removing the system response from measured UWB SAR images involves operations in the two-dimensional spatial frequency domain of the imagery. We show that correction can also be performed by convolving the polarimetric images with two-dimensional polarimetric calibration filters (PCF). When compared to the frequency domain procedure, filtering offers improved calibration on extended-area targets, lower sidelobe noise, and computational savings. PCF can be synthesized from knowledge of the system response, or by an adaptive equalization technique that uses clutter returns to generate filters to correct for leakage, or crosstalk, between channels.