Diagnostic within a dielectric micro-structure: time-reversal and sparsity-constrained imaging
Résumé
Computational imaging of micro-structured systems
as a grid-like set of a finite number of circular cylindrical
dielectric rods is considered. How to achieve meaningful localization,
resolution-enhanced or, even, super-resolved, vs. the
wavelength(s) of operation if defects or radiators are to be found
within is discussed here. That is, the set can be seen from all
around when either a line source inside an intact system of
rods is to be localized, or when, now illuminated by exterior
line sources, there is an unknown distribution of missing rods.
One-shot time-reversal (mostly reduced as time-harmonic backpropagation,
though cases involving broadband transient data
will be proposed in the presentation) and iterative sparsityconstrained
joint source imaging are analyzed as solution tools
in both TE and TM polarizations.