Moving target detection for synthetic aperture radar via shadow detection

In this paper a novel technique is presented for detecting moving targets in SAR imagery by tracking the shadow of the target over multiple looks instead of using the target direct energy return. Unlike the SAR image of a moving target that is blurred and displaced, the shadow of the target is displayed at the true target location. The detection performance using shadow is primarily independent of the target radar crosssection and depends upon the clutter to noise signal level. A change detection method is developed that provides detection via identification of the target shadow as an outlier from a background distribution. The background is inferred from a sequence of SAR images. It is shown that simple temporal averaging by alignment of the image for estimating the background works reasonably well with SAR images offering good shadow contrast. The technique was successfully demonstrated initially using real radar data from a short range spotlight SAR system. Results against weaker shadow signals at longer stand-off range have also proved very encouraging showing a moving target being detected at 10 km standoff range.