PHD filter for single extended target tracking
Xinglin Shen, Zhiyong Song, Hongqi Fan, Qiang Fu · 2016
In this paper, we study using probability hypothesis density (PHD) filter to track single extended target. In order to obtain the correct target state estimation of the target, we remove the outliers of the PHD filter results and average the remaining target state to get a target state estimation. In this case, both the presence of the target and the number of extended target measurements can be estimated. What's more, as processing the extended target as multiple target, PHD filter only need the likelihood function of the sensor, which avoid the high computational complexity of calculation of extended target function likelihood. The simulation results show that the PHD filter has good performance in single extended target tracking.