Development of a multiview time-domain imaging algorithm with a Fermat correction

An imaging algorithm is presented based on the standard assumption that the total scattered field can be separated into an elastic component with monopolelike dependence and an inertial component with dipolelike dependence. The resulting inversion generates two separate image maps corresponding to the monopole and dipole terms of the forward model. The complexity of imaging flaws and defects in layered elastic media is further compounded by the existence of high contrast gradients in either sound speed and/or density from layer to layer. To compensate for these gradients, we have incorporated Fermat’s method of least time into our forward model to determine the appropriate delays between individual source-receiver pairs. Preliminary numerical and experimental results are in good agreement with each other.