Application of motion reconstruction to ISAR data

Three approaches are applied to reconstruction of an object image from radar data collected in an Inverse-SAR (moving object, fixed radar location) geometry. These approaches attempt to reconstruct the object motion from the radar data alone. While all three approaches work with simulated data, application to this real-world data did not produce a recognizable image. Presented is a short description and comparison of the three methods, and what difficulties were found when applying these motion estimation algorithms to the real data set. The data that were used had been pre-processed into a time series of brightness-vs-range profiles. These range profiles were provided without any further information about the object or radar, providing a truly `blind' test.