This thesis presents an unified framework of inverse methods for the interpretation of time-domain electromagnetic (TDEM) data. The objective is the systematic and physically consistent extraction of geometric and material properties from transient EM responses. This is enabled by analytical models that establish a direct link between measured signals and the underlying properties of investigated objects.
First, an accurate TD radar-imaging technique for the shape reconstruction of multiple objects is developed. A detailed analysis of position-dependent imaging errors in multi-object configurations is performed. Based on this, an improved signal processing approach is proposed.
The work further addresses the inverse characterization of EM constitutive properties of metasurfaces. Two complementary approaches are considered. The first is based on an analytical TD reflectometry (TDR) technique. The second relies on a stochastic optimization procedure. Both are evaluated in terms of accuracy, robustness, and practical limitations.
Finally, the experimental realization of inverse interpretation is demonstrated. Electrical conductivity is estimated from the pulsed EM-field close-range coupling between loop antennas. The approach is validated in biological tissue analysis and subsurface imaging of buried conductors.
The results establish TD inverse methods as a robust and versatile framework for extracting physically meaningful properties from transient EM measurements.