Multiband Microwave Sensing For Surface Roughness Classification

This contribution presents an approach for frequency dependent surface roughness classification algorithms. For both simulations and measurements of the backscattering on random rough surfaces statistically defined surfaces were synthesised as well as manufactured. The coexistence of quasiidentical real and model surface profiles enables the validation of scattering models. The Fresnel-Kirchhoff diffraction theory will be applied and compared with a full-wave simulation software as a reference followed by measurements in the E and D band. Frequency dependency and roughness scaling effects will be worked out to derive a model-based methodology for the estimation of roughness depths as it is needed for applications like road condition detection.