Engebeli yüzey altına gömülü dielektrik cisimlerin üç boyutlu skaler durumda görüntülenmesi

Imaging of dielectric objects embedded in a layer, or beneath a rough surface is a popular subject of interest among remote sensing problems, and there are many studies that use microwaves for this purpose. In most of these studies, a rough surface separates two infinite half-spaces, and the object of interest lies in the inaccessible half-space. The area in which the object is presumed to be located is illuminated by transmitting antennas that are located in the accessible half-space, and the scattered field, which is the result of the interaction between the incident field and the irregularities in the medium, such as the unknown object and the layer it lies within, is measured by receiving antennas. The measured data is then used to obtain the unknown geometrical and material properties of the object. In the literature, the majority of the studies deal with layered media with planar interfaces, and 2D cases. Moreover, most of the studies on 3D cases dwell on only detection of dielectric objects, and very few deal with the imaging of dielectrics buried under a rough surface despite the importance of the problem from both theoretical and practical points of view. Nevertheless, none of these studies apply one of the well accepted nonlinear inversion techniques to a highly complicated case such as 3D case of imaging objects buried under a considerably rough surface, which is investigated here. In this thesis, a nonlinear tomographic approach for a 3D scalar case of microwave imaging of dielectric objects buried under a rough surface is presented. First, the imaging problem is reduced to a system of two integral equations, which requires the Green's function of the background medium, including both the two half-spaces, and the rough surface. Therefore, the Green's function of the background is obtained numerically by using Buried Object Approach (BOA), which involves obtaining the Green's function of the two half-spaces with a planar interface, and then treating the roughness as a series of objects embedded in both half-spaces alternately. Then, the system of integral equations is solved for the contrast function via Contrast Source Inversion (CSI) method, which is one of the most successful nonlinear inversion techniques when the Green's function of the background is known. The efficiency of the method is tested numerically, and successful results are achieved for different frequencies, permittivity and conductivity values of both the medium and the object, different heights of the roughness, different size and depth of the object, and for cases of multiple objects.