An Improved Multiple Hypothesis Tracker Integrated with Unknown Clutter Intensity Estimator

The multiple hypothesis tracker (MHT) is the optimal multi-target data association algorithm in ideal conditions. However, there is a degraded accuracy of data association when the observation scenario containing unknown clutter density. To addressed this problem, this paper proposed a new multiple hypothesis tracker integrated with unknown clutter intensity estimator (MHT-UCIE). In the MHT-UCIE, the clutter intensity is formulated as the product of the average number of clutter points per scan and the clutter spatial density, which are obtained subsequently through the maximum likelihood estimator and the Gaussian kernel estimator. Finally, the clutter intensity and track hypotheses are calculated. The experiment results verify the superior tracking performance of the MHT- UCIE in an unknown clutter intensity tracking scenario.