Improved General Polarimetric Model-Based Decomposition for Coherency Matrix

A representative general polarimetric model-based decomposition framework was proposed by Chen et al., which implements a simultaneous full-parameter inversion using complete information and solves several limitations in previous methods. However, there are still shortcomings Chen’s work. Firstly, only the real part of parameter β generalized surface scattering model is considered. Secondly, inappropriate initial input values may lead to local optima nonlinear least squares optimization algorithm. Thirdly, volume component underestimated scattering-dominated scene, but overestimated buildings with large orientation (LOB) areas. Finally, time-consuming computationally. To overcome those issues, an improved method this paper. The imaginary incorporated into method. Ingeniously utilizing internal relationship generic equations composed coherent matrix elements, parameters can be inversed simplifying linear equations. Therefore, compared method, does not rely on values, improves computational efficiency. In addition, hierarchical scheme presented solve problem underestimation or overestimation mentioned above. performance advantages evaluated L-band C-band synthetic aperture radar (PolSAR) data sets. Comparison studies carried out other methods, demonstrating that further improve performance, especially LOB