Computational Imaging Using Frequency-Diverse Scattering from a Rough Reflector

This paper introduces a random rough reflector with walls to produce frequency uncorrelated scattering to image an object. The conducting rough surface, surrounded by 4 walls, resembles an open-ended cavity. First, the scattered fields are computed using the geometrical optics (GO) approximation. Next, the least-squares computational imaging algorithm is implemented and numerical images are computed using 401 frequency samples within the frequency range 2-20 GHz. Finally, the imaging setup is made and measurement is conducted. Both simulated and measured data reveal a clear image of the target. For a one square meter (m²) image, a descent pixel size of about 1/401 m² (5 cm x 5 cm) is obtained.