Premium
Analysis of the Super‐Resolution Effect on Microwave Tomography
Author(s) -
Simonov Nikolai,
Jeon SoonIk,
Kim BoRa,
Son SeongHo
Publication year - 2018
Publication title -
radio science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.371
H-Index - 84
eISSN - 1944-799X
pISSN - 0048-6604
DOI - 10.1029/2017rs006404
Subject(s) - singular value decomposition , truncation (statistics) , microwave imaging , inverse scattering problem , inverse problem , singular value , noise (video) , nonlinear system , tomography , microwave , scattering , iterative reconstruction , resolution (logic) , inverse , image resolution , algorithm , mathematics , optics , computer science , physics , mathematical analysis , image (mathematics) , eigenvalues and eigenvectors , geometry , statistics , telecommunications , artificial intelligence , quantum mechanics
This study investigates the spatial resolution (SR) and super‐resolution effect in microwave tomography in terms of single‐frequency measured data. The applied method is based on our recently proposed concept of average SR (ASR). We apply truncated singular value decomposition to calculate a regularized forward‐modeling matrix and to limit the truncation index by the acceptable level of the imaging noise. A simple relation of the ASR with the truncation index calculates the ASR in the imaging zone. The described method to calculate SR is quite common, and it considers not only the geometrical parameters of the microwave tomography system and object under test but also the noise in the measured data. This method is applicable to the linear and nonlinear considerations of the inverse scattering problem with respect to two‐ or three‐dimensional solutions. In particular, our investigation confirms the conclusion of some other authors that applying nonlinear inverse scattering methods can achieve the super‐resolution imaging even when based on far‐field measured signals only.