Develop low-cost high-speed image processing and visualization techniques for persistent surveillance applications.

Persistent surveillance sensing allows for the production of synthetic aperture radar (SAR) movies that promise to allow continuous tracking of moving targets on the ground in dense traffic environments and through move-stop-move cycles that confound traditional tracking approaches. Real-time production of a sequence of very large SAR images is a significant challenge that facilitates many desirable exploitation approaches. Current approaches rely on a digital spotlighting followed by a computationally expensive back-projection at their core. Our approach will decrease the computation cost of the digital spotlighting and replace the back projection with a modification of the popular polar format algorithm. Initial calculations indicate that our approach will reduce the number of processors by as much as an order of magnitude. We anticipate great interest from a number of radar manufactures in the licensing of this faster implementation of a persistent surveillance SAR mode. BENEFIT: This research will allow real-time production of very large synthetic aperture radar images from an orbital persistent surveillance geometry. The algorithm will improve image quality and reduce computational resources required for any existing airborne radar system. These existing radar systems may be used for anything from terrain mapping and border patrol to site monitoring for counter terrorism.