Imaging of subsurface regions of random media by remote sensing

The ability to selectively probe a random medium, even in the limit of isotropic scattering, suggests it should be possible to reconstruct images of a dense scattering medium from the information contained in the backscatter surface emission profile. Consideration of the imaging problem also requires knowledge of the impact that localized absorption at any site in the medium would have on the response of detectors at the surface. By applying the concept of importance used in reactor theory, we have calculated this relationship for various homogeneous media. This information was subsequently incorporated into several image reconstruction algorithms which employ a backprojection strategy. The algorithms were tested by applying them to simulated surface emission profile data for homogeneous scattering media with embedded arrays of black-body absorbers. The algorithms correctly identified the size and location of the arrays, resolved internal structural features, and showed significant improvement upon iteration.