Combination of H-Alpha Decomposition and Migration for Enhancing Subsurface Target Classification of GPR
Xuan Feng,Yue Yu,Cai Liu,Michael Fehler
DOI: https://doi.org/10.1109/tgrs.2015.2411572
IF: 8.2
2015-01-01
IEEE Transactions on Geoscience and Remote Sensing
Abstract:Polarimetric technology has been one of the most important advances in microwave remote sensing during recent decades. H-alpha decomposition, which is a type of polarimetric analysis technique, has been common for terrain and land-use classification in polarimetric synthetic aperture radar. However, the technique has been less common in the ground penetrating radar (GPR) community. In this paper, we apply the H-alpha decomposition to analyze the surface GPR data to obtain polarimetric attributes for subsurface target classification. Also, by combining H-alpha decomposition and migration, we can obtain a subsurface H-alpha color-coded reconstructed target image, from which we can use both the polarimetric attributes and the geometrical features of the subsurface targets to enhance the ability of subsurface target classification of surface GPR. A 3-D full polarimetric GPR data set was acquired in a laboratory experiment, in which four targets, a scatterer with many branches, a ball, a plate, and a dihedral scatter, were buried in dry sand under flat ground surface, and used to test these techniques. As results, we obtained the subsurface H-alpha distribution and classified the subsurface targets. Also, we derived a subsurface H-alpha color-coded reconstructed target image and identified all four targets in the laboratory experiment.