Underlying Topography Inversion Using TomoSAR Based on Non-Local Means for an L-Band Airborne Dataset

The underlying topography is an important part of the three-dimensional structure forests, and used for a variety applications, such as hydrology water resource management, civil engineering projects, forest surveying. Due to imaging ability strong penetration, tomographic synthetic aperture radar (TomoSAR) with long wavelength has been shown be useful tool estimate topography. At present, most current methods use local means method sample covariance matrix, in which vertical backscattering power estimated. However, these cannot easily obtain high-precision topography, often lose some detailed information. In this paper, solve problem, non-local introduced optimal matrix by combining weighted neighborhood pixels. To validate feasibility effectiveness proposed method, BioSAR 2008 campaign L-band dataset acquired from northern forests Sweden was inverse results show that accuracy retrieved improved more than 30% when compared traditional method.