Linearized inversion methods for three-dimensional electromagnetic imaging in the multiple scattering regime

Abstract : In optical or microwave computational tomography, the sample permittivity is reconstructed numerically from the measurements of its scattered field for various illuminations. When the light sample interaction involves multiple scattering, the relationship between the scattered field and the permittivity is non-linear and a direct reconstruction is not possible. Using a simple physical approach, adapted to the three-dimensional vectorial electromagnetic framework, we derive an iterative inversion technique, based on the linearization of the scattering operator, for imaging (possibly anisotropic) targets in the multiple scattering regime. We investigate the performances of different approximations of this operator accounting for more or less multiple scattering. Our method is applied to the reconstruction of targets in the microwave domain using experimental data. ARTICLE HISTORY
Document type :
Journal articles
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02023456
Contributor : Guillaume Maire <>
Submitted on : Tuesday, February 19, 2019 - 4:16:32 PM
Last modification on : Monday, March 4, 2019 - 2:04:24 PM

Identifiers

Collections

Citation

Kevin Unger, Ting Zhang, Patrick Chaumet, Anne Sentenac, Kamal Belkebir. Linearized inversion methods for three-dimensional electromagnetic imaging in the multiple scattering regime. Journal of Modern Optics, Taylor & Francis, 2018, 65 (15), pp.1787-1792. ⟨10.1080/09500340.2018.1459912⟩. ⟨hal-02023456⟩

Share

Metrics

Record views

43