Time Reversal (TR) imaging methods have recently generated considerable interest due to their robustness and ability to locate multiple targets. However, problems arise when the targets are behind an obstructing material. We develop a preprocessing algorithm for TR imaging methods to obtain images of targets hidden behind a wall, applying temporal and spatial windows on the full-Multistatic Data Matrix (MDM). By using temporal-windows, we divide the full-MDM into sub-MDMs. We further segment the sub-MDMs spatially by selecting its elements corresponding to a predetermined number of antennas in space.
We take the singular value decomposition of the sub-MDMs to obtain singular value distributions as well as singular vectors which we analyse individually and use for the MUltiple SIgnal Classification (MUSIC) TR imaging method. The obtained images from the these localised sub-MDMs collectively yield to clearer
images of scenarios than we can obtain using the full-MDM.