Direct Trajectory Estimation of Non-Linear Moving Objects

Inhalt:
The accuracy of the trajectory and velocity estimation of moving objects, like cars and ships, in synthetic aperture radar (SAR) images using ground moving target indication (GMTI) techniques is limited. Most algorithm estimate a constant velocity over a short coherent processing interval (CPI), despite the fact that object motion is in general non-linear. We formulate a signal model including non-linear object motion and propose a direct trajectory estimation algorithm. Our algorithm uses the robust displayed phase center antenna (DPCA) technique to suppress the clutter. It basically compares two DPCA images to estimate a smooth object motion by a regularization method. We apply the technique to simulated data to show that the estimation of a non-linear movement of an object with a certain size can be done by the proposed algorithm.