Session 3A4 Microwave Imaging for NDE/NDT Applications

In this paper, a neural network approach is applied to a typical electromagnetic diagnostic problemconsisting in the prediction of the electromagnetic field absorption inside a dielectric phantom. The approachhas been tested by considering, at the input of the neural network, the values of the incident electric field ata fixed number of locations. Also phaseless measures have been taken into account. The outputs are somefeatures describing the electromagnetic absorption, such as the peak amplitude of the induced field, the locationof the absorption peak and a measure of the extension of the induced spot inside the phantom.Preliminary results show that the approach shows a satisfactory accuracy in reconstructing the selectedabsorption features, and that it is able to estimate these characteristics very quickly.