Modified Entropy-Based Fully Polarimetric Target Classification Method for Ground Penetrating Radars (GPR)

Abstract

Polarimetric scattering characteristics are useful for classifying different radar targets. In particular, the entropy-based H-α classification method has been widely used in remote sensing for classifying land features with different polarimetric scattering properties. However, this method has not been widely adopted in ground penetrating radars (GPR) applications due to the lack of commercial fully polarimetric GPR systems. This paper will demonstrate that the conventional nine-zone H-α classification chart adopted in remote sensing needs to be modified in order to be effective in classifying GPR targets due to the common presence of noise and clutter. The effects of white noise and clutter on H-α features derived from the fully polarimetric scattering data are discussed. This study led to the development of a new four-zone H-α classification chart. The application and effectiveness of classifying linearly polarized and nonpolarized targets using this new chart will be demonstrated via outdoor measurement examples.