Multiple Target Tracking


Multi-target tracking (MTT) becomes a compulsory task in modern surveillance system where a large amount of information about the targets is collected, communicated, combined. The main objectives in the MTT system are to correctly identify the targets and to well maintain the tracks by accurate track update. This report describes such a tracking method for multiple targets in background clutter via a real radar data based simulation. The main component of the MTT system such as track initiation, data association and track update, gating computation, etc. are introduced. In particular, in the simulation study, we employ the logic-based method for track initiation, Munkres algorithm for data association, and Kalman filter and interacting multiple model (IMM) filter for both maneuvering and non-maneuvering target tracking, respectively. Simulation study used the experimental data collected from ANIFPN-504 air traffic control radar and the ANIFPS-SO8 air defense radar while the flying courses were taken place at Canadian Forces Base. Both maneuvering and non-maneuvering targets can be tracked. Maneuvering target tracking demo and non-maneuvering target tracking demos are resided in the Download section.