The fast linear multisensor RFS-multitarget tracking filters
Weifeng Liu, Chenglin Wen · International Conference on Information Fusion · 2014
The main filters in the finite set statistics (FISST) framework include the PHD filter, the CPHD filter and the MeMBer/CBMeMBer filters. We referred all theses filters as the RFS-multitarget filters. This paper mainly deals with the multisensor with the property of linear correlation for the RFS-multitarget filters in the product space of multisensor. We proposed the linear Multisensor RFS-multitarget (LM-RFSM) filters by using the measurement dimension extension (MDE) approach, which remains the same appearance like the conventional RFS-multitarget filters except the produce space and some parameters in the filters. In the product space of sensors, the dimension extended measurements may greatly increase the computational load. In order to improve the computational speed, we propose a fast algorithm for the LM-RFSM filters. The experiment shows that the fast algorithm can greatly increase the running and at the same time the impact on the performance is small.