Ripple phenomenon and artifact elimination in millimeter-wave imaging of a hidden target

Abstract

In millimeter-wave imaging of a hidden target, the effect of the dielectric cover before the target is typically ignored. This results in ripple-corrupted images that pose challenges for target recognition. In this paper, we provide a perspective for understanding the image of the hidden target, which clearly reveals the origin of the ripples, and propose a separation method that not only gets rid of ripples, but also obtains the target’s depth map. Reflections and transmissions during imaging are considered and decoupled to separately form images corresponding to each real or virtual object. An algorithm based on the range-direction spread function is developed to iteratively estimate the depth and reflectivity of the target. Imaging experiments with and without a cover are conducted to demonstrate the formation and influence of ripples and to verify the proposed algorithm. Our work deepens the comprehension of covered target imaging. Benefited fields might include non-destructive testing, through-wall imaging, subsurface imaging, and security screening.