Bifocusing-Based Microwave Imaging Under Background Conductivity Mismatch

Abstract

Microwave imaging is a promising technique for non-invasive diagnostics in areas such as medical imaging, remote sensing, and subsurface exploration. Its performance, however, strongly depends on accurate knowledge of background medium properties, particularly electrical conductivity. In practice, these parameters are often uncertain or mismatched, leading to signal degradation and inaccurate reconstructions. This study investigates the impact of conductivity mismatch on the bifocusing method (BFM), a qualitative microwave imaging algorithm, and proposes an improved version incorporating attenuation compensation. Simulations show that conventional BFM fails to identify objects when the assumed conductivity is significantly higher than the actual value. To resolve this, we modify the Green's function by introducing a compensation term based on the attenuation constant, restoring the incident field amplitude. The improved method enables successful object recovery even under severe mismatch. Quantitative evaluation using the Jaccard similarity index confirms improved localization accuracy. This approach enhances the robustness of microwave imaging and shows promise for medical diagnostics in highly attenuating biological tissues.