The polarimetric matched image filter: application to speckle reduction and optimal background clutter discrimination in microwave sensing and imaging

摘要

Summary form only given. Scattering matrix optimization procedures were applied to the complex scattering matrix, associated covariance matrix, and power density matrix and their Lie SU(n)-group theoretic expansions (n=2,3,4), in order to derive a set of more robust target detection, speckle reduction, and clutter suppression algorithms. The covariance matrix invariance ratio was introduced and found to provide a good measure of speckle reduction as well as being useful for obtaining an optimal high resolution image. A PMSF/PMIF (polarimetric matched signal filter/image filter) filtering process, which is based on the target/clutter characteristic polarization state plus optimal stochasticity coefficient approaches, was developed. Results were obtained for POL-SAR image data sets collected with the airborne NASA-JPL (P/L/X)-band SAR polarimetric system over the San Francisco Bay area.