Microwave imaging of hidden targets in a complex scattering medium has drawn much attention as it can be used to gather information of concealed targets. Among these, through-wall radar imaging (TWI) is an emerging technology for “seeing” through walls to determine building layout and occupancy. This dissertation focuses on the development of special signal processing techniques to address the impact of wall distortion on the interior image. The goal is to develop radar imaging techniques that incorporate electromagnetic propagation models of the wall structure to improve target restoration.
This dissertation begins by establishing an understanding of the fundamental synthetic aperture radar (SAR) imaging principles along with its model-based extension to mitigate the wall reflection and propagation delays of the uniform dielectric walls (multiple layers). Specifically, wall compensation is carried out via an Adaptive CLEAN (A-CLEAN) and target refocusing algorithm. Subsequently, periodic wall structures are characterized using Floquet modal analysis and plane wave spectral expansion (PWE). To further improve the computational efficiency, a high-frequency ray model that approximates the exact solution by a set of rays is also presented. It is shown that the structural periodicity induces higher-order space harmonics leading to risen clutter and
ghost artifacts in the through-wall image.
To overcome these distortions, this dissertation presents a model-corrected inverse imaging framework that incorporates the periodic layer Green’s function into its forward model. For that, a linear back-projection solution and a nonlinear minimization solution are applied to solve the inverse problem. The back-projection image corrects the distortion and has higher resolution compared with free space due to inclusion of multi-path propagation through the periodic wall, but considerable sidelobe clutter is also present. On the other hand, the nonlinear solution not only corrects target distortion without clutter, but also reduces the solution to a sparse form.
A multi-path imaging approach is also proposed to exploit the multi-scattering effects more directly to our advantage. Specifically, the imaging kernel of the back-projection method is designed to focus any propagation paths of interest. Subsequently, an adaptive sidelobe reduction technique based on spatially-variant apodization (SVA) is applied to suppress the unwanted sidelobes. It is shown that the Floquet modes of the periodic structure greatly increase the effective radar aperture leading to much improved target resolution than that in free space. The research findings can also be attractive to other microwave imaging applications.