Signal Processing And Application Of Polarimetric SAR/InSAR

Polarimetric Synthetic aperture radar (PolSAR) can obtain the polarimetricscattering matrices of targets in the scene via adjusting the transmitting and receivingantennas polarizations. Polarimetric interferometric SAR (PolInSAR) possesses theabilities of both PolSAR and InSAR, which can overcome the problems that exist in themeasurements of InSAR by using the polarimetric information so that improves theaccuracy of measurement. Therefore, PolSAR/InSAR has become one of the mostimportant developing directions of SAR systems.With regard to the processing difficulties and hot applications of PolSAR/InSAR,in this thesis, PolSAR/InSAR signal processing and applications are researched on.In Chapter two, the principles of PolSAR and PolInSAR are illustrated, includingthe polarization of electromagnetic wave, the polarimetric scattering matrix, thescattering mechanism, the complex polarimetric interferometric coherence, etc.In Chapter three, PolSAR/InSAR pre-processing, including polarimetric calibrationand polarimetric filtering method, are researched on. Firstly, the basic calibrationprinciples and some calibration algorithms are presented and validated by simulateddata and real data taken by the vehicle-borne X-band PolInSAR system, which belongsto National Laboratory of Radar Signal Processing, Xidian University. Secondly, anovel method is proposed to reduce speckles and preserve the scatter characteristicssimultaneously. The comparison with the classical Refined Lee Filter shows theeffectiveness of the proposed method.In Chapter four, some representative applications of PolSAR/InSAR are studied,including PolSAR classification, PolInSAR classification and PolInSAR targetdetection beneath foliage. Firstly, two unsupervised PolSAR classification methods aremodified to improve their performance. Secondly, a polarimetric database is built withthe data taken by the vehicle-borne system, with which a supervised classification isconducted. Thirdly, a PolInSAR classification method with high accuracy is proposed.Compared with other algorithms, it can classify images directly with the polarimetricand interferometric information, avoiding of PolSAR classification. Finally, the classicalCloude algorithm is analyzed and a beneath foliage target detection method is proposed.The experiment results show that the new algorithm can reduce the computation timedramatically without the loss of accuracy. Finally, some aspects which need improving are pointed out.