Model-Based Polarimetric Target Decomposition With Power Redistribution for Urban Areas

• Published in: IEEE journal of selected topics in applied earth observations and remote sensing Vol. 16; pp. 8795 - 8808
• Main Authors: Hu, Canbin, Wang, Yifei, Sun, Xiaokun, Quan, Sinong, Xiang, Deliang
• Format: Journal Article
• Language: English
• Published: Piscataway IEEE 2023; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Buildings; C band; Coherence; Cross-polarized scattering; Decomposition; Electromagnetic scattering; Matrix decomposition; oriented built-up areas; overestimation of volume scattering (OVS); polarimetric decomposition; Polarimetry; power redistribution; Power transfer; Radarsat; Scattering; Solid modeling; Urban areas
• ISSN: 1939-1404; 2151-1535
• DOI: 10.1109/JSTARS.2023.3314129

Polarimetric decomposition of oriented buildings is challenging due to their variable orientation angles and structures. Both vegetated and oriented built-up areas generate the HV component namely cross-polarized scattering, leading to an overestimation of volume scattering (OVS). It can cause misinterpretation of the scattering mechanisms between oriented built-up and vegetated areas. In this article, we use a pure volume scattering model designed for completely random scattering to describe the scattering of vegetated areas. At the same time, a new urban revised rate is proposed by considering the rotated dihedral model and introducing polarimetric asymmetry, which can distinguish different areas and reduce OVS in urban areas through power transfer strategy. Then, a new model-based polarimetric target decomposition with power redistribution for urban areas is proposed. The performance of our proposed method is verified by RADARSAT-2 C-band and UAVSAR L-band data. The results show that our method can not only better characterize the scattering of oriented buildings compared to previous methods but also maintain the scattering power of natural areas. It can alleviate the scattering misinterpretation problem between oriented built-up and natural areas.