Compact polarimetric synthetic aperture radar (SAR) is a dual-polarimetric system, which shows promise of reducing the complexity, cost, mass, and data rate of a SAR system. It allows fully polarimetric information reconstructed from compact polarimetry (CP) under some constrains. Two assumptions are required here for the constructing, one is the reflection symmetry, and the other is a relationship model between the co-polarized and cross-polarized channels. In this paper, an improved reconstruction algorithm is proposed according to both the assumptions, and two improvements are according proposed. Since the helix scattering component is always used to take account of the non-reflection symmetric circumstance which mostly appears in complex urban area, helix scattering should be considered in the reconstruction algorithm, especially for the regions with strong orientation angle effects. A modified four-component model-based decomposition algorithm is proposed with a new volume scattering model. Then using the decomposed powers of the four components, the relationship between the co-pol and cross-pol channels is developed instead of the original polarization state extrapolation model. Finally, a ESAR L-band polarimetric dataset over Oberfaffenhofen area is used to demonstrate the effectiveness of the proposed algorithm.
展开▼