Advance of satellite-borne microwave remote sensing programs, such as active SAR, ERS scatterometer etc., during recent decades has promoted study of theoretical modeling, data calibration/validation, image analysis, surface classification, andparameter inversion for these programs{sup}([1-2]) In this paper, to simulate fully polarimetric scattering from tree canopy applicable to SAR imaging at C band, the scattering amplitude functions of randomly oriented, non-spherical particles (e.g. leaves with thin thickness, branches with small radius) are derived in the generalized Rayleign-Gans (GRG) approximation. Numerical results show the difference between the GRG and conventional Rayleigh approximations. These functions averaged over randomlyspatial orientation are employed to construct the non-diagonal extinction matrix and the Mueller matrix solution. Four Stokes parameters of scattered fields are coupled. It is different from conventional iterative solution of isotropic scattering media.Further, to physically identify polarimetric scattering of the Mueller matrix solution, the coherency matrix and its eigen-analysis are discussed. Information of the eigenvalues and entropy from the Mueller matrix measurement is related with polarimetricscattering process and medium randomness, and can be used to investigate directly the influence of different scattering mechanism in complex scenarios on observable parameters. Asan application, the eigenvalues and entropy derived from the Mueller measurement of an AirSAR image are demonstrated.
展开▼