Ship Detection and Measurement of Ship Motion by Multi-Aperture Synthetic Aperture Radar
Defence Research and Development Canada Ottawa Ottawa, Ontario Canada
Pagination or Media Count:
SAR Synthetic Aperture Radar GMTI Ground Moving Target Indication studies of ship signatures over a ten year period have developed an adaptive approach to the creation of ensembles of adaptive range-history models that characterize ship motion, ship length and ship heading from multi-aperture airborne and space-based radar measurements. The ships being detected and measured are assumed to be rigid bodies that move in response to propulsion and sea motion forces. Non-parametric moving target detection and motion estimation metrics are applied to radar returns from ship targets to represent each observed ship as an ensemble of spatially clustered moving targets whose motions, clustered into domains over the ship target area, estimated metrics are fed back to range-history models to improve the SAR focusing of the ship and to improve the motion estimation accuracy. DPCA Displaced Phase Center Antennamoving-target detection algorithms were used to generate lists of moving-target candidate points from two and four aperture SAR-GMTI scenes and the signal data corresponding to these points were analyzed to cancel sampling ambiguities from ships and littoral zone land targets prior to target motion estimation for each sample point. At the end of the estimation process target samples were clustered to extract the ensemble bulk properties of the ship target. When highly over-sampled GMTI data are acquired from airborne radars, coherent processing intervals that correspond to ocean wave periods that influence ship motion are obtained and ocean motion estimates can be extracted from the ship data using time-frequency analysis. The incorporation of the ocean motion estimates into the range history set that represents the ship allows fine SAR focusing of the ship image. An example is shown for a 167 m ship transiting off the coast of Newfoundland.