摘要

When traditional single-point target tracking algorithms are directly applied to multiplepoint rigid targets, filter divergence and data association problems occur. By introducing the concepts of global motion, local motion and structure parameters, and employing motion-induced invariants on local motion and structure parameters, we have developed several kinematic and measurement models to simultaneously estimate the motion and structure information of a 3D rigid-body target from high range resolution (HRR) and ground moving target indicator (GMTI) measurements. To test the ability of the proposed modelling methods in handling false alarms, occlusions and HRR measurements without correspondence information, several data association approaches are also proposed, that is permutation, comb ination-perin utation, dummy measurements, and dynamically changing the interested number of scattering centres. Simulation results have shown that the proposed modelling methods can successfully estimate both the motion and structure information of an object by using HRR and GMTI measurements with a proper data association method.

  • 出版日期2007-7