ISAR target parameter estimation with application for automatic target recognition

Algorithms for estimating the angular velocity, angular acceleration, orientation, length, and width of an ISAR target and for identifying the scattering points of ISAR targets are derived and demonstrated. A correspondence between the ISAR motion estimation operations and a rigid body dynamical model is illustrated. It will be shown that an ISAR image amounts to projecting scatterers into a plane perpendicular to the target axis of rotation and coplanar with the range axis. The proper Doppler scale factor of imagery and its variation is proportional to the rotation rate and acceleration. These result as a by-product of SAIC ISAR motion estimation techniques. Improved geometric estimates can then be found by incorporating modeling assumptions and a priori knowledge. A number of land and sea examples are provided to reinforce the above concepts.