Adaptative Stolt migration via contrast maximization for GPR applications

The Stolt migration, also referred as frequency-wavenumber migration, is a Fourier transform based processing technique to focus GPR scattering objects back into their true physical shape and position. This procedure requires a priori knowledge of the velocity distribution of the subsurface (i.e. the soil dielectric permittity) and this information is usually not available and needs to be retrieved from surface measurements with for example a TDR probe. In this paper we propose a simple but efficient way of iteratively estimating the background permittivity based on the maximization of the energy contrast when migrating a GPR image.