Development of a Novel Method for Through-Wall Imaging

摘要

Through-wall imaging (TWI) consists in imaging objects hidden behind an obstacle by using electromagnetic waves at microwave frequencies. Developing systems which 'see' through walls gives the ability to detect and localize people, observe room layouts and understand scenarios that are hidden. The crucial point to be addressed is the presence of the obstacle representing in some cases a wall structure. Indeed, the presence of the wall, on one side, can hinder a successful imaging procedure due the absorption and to the refraction losses it introduces. However, for frequencies ranging from a few hundreds of MHz to 2-3GHz most of the building materials are relatively transparent; which makes high-resolution TWI feasible. On the other side, the presence of the wall has to be embodied in the imaging procedure otherwise blurred images are obtained. In this proposal we aim at developing a novel TWI procedure based on the microwave tomographic approach and to validate it against synthetic and experimental data collected in realistic conditions. Accordingly, the TWI is cast as an inverse scattering problem where the presence of the wall is straightforwardly (with respect to other methods) embodied within the so called background Green's function. More in detail, despite the inverse scattering problem is non-linear, we adopt a linear reconstruction algorithm based on the Kirchhoff approximation as it allows to meet all the main requirements for a TWI. First, it generally allows to determine the scatterers' locations and their geometrical features well beyond the limits of the approximate model. Moreover, it allows to obtain reliable reconstructions free from the local minima problem and is computationally effective, permitting thus to deal with large (in terms of wavelength) investigation domain in almost real time. Finally, well assessed numerical and theoretical tools exist to achieve stability against noise.