Gaussian Mixture PHD Filter with Measurement-labelled Adaptive Target Birth Intensity
Jihong Zheng, Meiguo Gao, Haojie Yu · 2018
The Gaussian mixture probability hypothesis density (GMPHD) filter is an efficient and real-time method for estimating multiple target states with varying target number in clutter. However, the main drawback of this method is that the target birth intensity is known as a priori. In other words, the GMPHD filter is inapplicable to the tracking scenario with no priori spatial information on where targets can appear. To address this limitation, a measurement-labelled adaptive target birth intensity for GMPHD filter is proposed in this paper. All key equations of this proposed method are derived, and a multi-target tracking scenario is designed to demonstrate the performance of the proposed method. Simulation results suggest that the measurement-labelled method is an effective and efficient solution to target birth intensity adaptive problem.