Multiple Radar Targets Tracking and Trajectories Fitting

This paper proposes a data association and tracking algorithm followed by polynomial fitting, in the case of multiple radar targets to improve the accuracy of targets’ trajectories estimated by an automotive radar sensor.Starting from the conjecture that any real trajectory is polynomial in cartesian coordinates the proposed method supposes to fit the data obtained at the output of Kalman filter to a polynomial. We used the polynomial fitting method for the trajectories of three targets, which we generated in two different scenarios. In both scenarios the targets could be tracked properly using Kalman filters after data association, and the trajectories were improved by applying the polynomial fitting.