Urban terrain multiple target tracking using probability hypothesis density particle filtering

A multi-model particle probability hypothesis density filer (PPHDF) algorithm for multiple target tracking in urban terrain is investigated in this paper. The multi-model PPHDF is based on target state-space modeling of urban scenarios, random finite set theory, multiple model estimation theory, and sequential Monte Carlo implementations. Our proposed algorithm can instantaneously and efficiently estimate both the number of targets and their corresponding states without conventional measurement-to-track associations. Numerical simulation results demonstrate that the multi-model PPHDF can achieve good tracking performance with tractable computational complexity in the test bench urban tracking scenario with complex multipath radar return patterns.