Comparison of PHD based filters for the tracking of 3D aerial and naval scenarios

The Probability Hypothesis Density (PHD) filter is applied to realistic three-dimensional aerial and naval scenarios to illustrate its performance in detecting, initiating and terminating tracks in presence of clutter. Radar measurements are available every two seconds. A comparisons between different approximations of the PHD recursion, namely the sequential Monte Carlo and the Gaussian Mixture approximation, is given on different scenarios using the OSPA metric and different levels of clutter.