Polarimetric enhancement of pol-SAR imagery applied to JPL-AIRSAR polarimetric image data

Radar polarimetry (the full utilization of EM vector wave information) has become an indispensable tool in terrain imaging technology. Polarimetric imaging (polarization filtering) has been carried out by many investigators. Boemer´s group developed the polarimetric-filtering principle for both coherent and incoherent cases using the polarization ratio and Stokes vector formulations. The CAL-TECH/JPL(NASA) group applied the principle based on the Stokes vector formulation to the imaging of polarimetric SAR data acquired at NASA JPL. Touzi et al. proposed a filtering technique for SAR images for reducing speckle. Swartz et al. and Boerner et al. solved the problem by using a decision-theoretic covariance matrix approach. Although many other approaches are available for polarization filtering, the distinct radar channel formulation is used, i.e., either the matched, the co-or cross (X)-polarization channel. This paper applies the principle to SAR imagery for these three possible polarimetric radar channels and compares the resultant imagery. The radar channels considered are co-pol, x-pol, and matched-pol. channels