|
ABSTRACT -- A penetrative imaging of 2D or 3D objects embedded
in planar multilayer media is investigated under a
multifrequency plane wave illumination. Both theoretical and
experimental results are presented in this dissertation. A 3D
multilayer Green''s function is derived in terms of plane wave
superposition by using a 2D Fourier transform (FT). Based on
this multilayer Green''s function, a penetrative imaging of 3D
embedded objects can be expressed as multifrequency plane wave
backward propagation. The 3D penetrative image can be
reconstructed by a 2D inverse Fourier transform (IFT) on stacks
of parallel planes, and the 2D penetrative imaging can be
reconstructed by a 1D IFT on stacks of parallel lines. If the
embedded objects are in the last layer, the penetrative imaging
can be implemented with a 2D or 3D IFT for 2D and 3D cases,
respectively. The penetrative imaging of both embedded
conductors and dielectrics are demonstrated. Moreover, the
performance of penetrative imaging with the random noise and the
interference is discussed. An automatic measurement system is
implemented for the penetrative imaging. Experimental results
under various multilayer media and embedded objects are
presented to demonstrate the performance of the imaging
algorithm. The numerical and experimental results reveal that
the proposed penetrative imaging can be applied to both the 2D
and 3D embedded objects whether they are conductors or
dielectrics. Besides, the method is insensitive to the high
random noise and the high interference up to 100% of the peak of
scattered field. Further developments into practical
applications in non-destructive testing and underground
detection are foreseen.
|