A hybrid parametric, non-parametric approach to Bayesian target tracking

This article describes a versatile approach to nonlinear, nonGaussian noise target tracking which makes use of both parametric and nonparametric techniques within a Bayesian framework. It produces a Gaussian mixture model (GMM) of a track, but resorts to a sampling technique within the tracking process to handle nonlinearity. GMMs are recovered from samples using the expectation-maximisation method. The approach has been implemented in PV-WAVE software and tested against a Kalman-filter tracker in a simulator with air-defence scenarios. Sample results are presented for a scenario with a single surveillance-radar and a single target following a weaving path. These show that the tracker produces significantly better position estimates and comparable heading and speed estimates. Computation times are about 30 times greater than for the Kalman-filter tracker, but there is scope for reducing that substantially by tolerating fewer samples