In this paper we investigate the robustness and the effectiveness of a microwave imaging technique, based on Bayesian estimation theory, for the reconstruction of dielectric profiles. The validation is conducted on real experimental data, the well-known "Marseille" dataset. Our statistical based inversion algorithm takes advantage of Bayesian regularization, which permits to invert a strongly non-linear model using a Markov Random Field as a-priori statistical model of the unknown image. Such choice leads to a robust and effective non-linear inversion method. An exhaustive analysis on the experimental data is also performed, in order to show the good performance of the method. © 2007 EURASIP.
|Titolo:||Bayesian regularization in non-linear imaging: Reconstructions from experimental data in microwave tomography|
|Autori interni:||AUTIERI, ROBERTA|
|Data di pubblicazione:||2007|
|Rivista:||EUROPEAN SIGNAL PROCESSING CONFERENCE|
|Appare nelle tipologie:||4.1 Contributo in Atti di convegno|