Ship and Boat Classification via 3-D Measurements on SAR-ISAR Data

MARK Resources has previously developed ISAR processing algorithms that provide undistorted plan and profile views of small ships and boats, and that associate responses in the two types of image. We propose to combine this processing with measurements on distorted plan and profile views and thereby derive the unknown crossrange scale of the undistorted profile view, and to extract target shape and dimensions from the resulting set of three-dimensional scatterer positions. We also propose to adapt the existing algorithms in order to obtain hybrid views with usable Doppler resolution for all three of the bow, stern, and superstructure; to measure response positions and associate responses in such hybrid images; and to derive three-dimensional positions of scatterers from the associated measurements. For both of the approaches, we propose to demonstrate the feasibility of generating a three-dimensional image of a small ship or boat and classifying it in real time. The demonstration will be carried out on data from the Navy"s SCATR collection.