Abstract

Inverse scattering and imaging go hand in hand and have multiple applications: medical imaging, geophysical prospecting or non-destructive testing. More recently, imaging, in the context of quality control in primary industry, has been used to understand the interior of other organic bodies: fruits or trees. There is still a huge demand from industrials to come up with a technology that allows to know the level of ripeness, sweetness, disease or any suitable properties of a product without cutting it open.     The fundamental task in imaging is indeed to infer the characteristics of a medium from indirect measurements. We developed a novel technique for this task called Adaptive Eigenspace Inversion (AEI). The AEI method has been shown to reduce computation time significantly and to introduce inherent regularisation, by projecting the domain into the basis of an elliptical operator during the implementation of Newton-type optimisation methods.     In this talk, I will present the AEI method in a first instance and then give some preliminary results in the context of microwave imaging on trees.