Multi-target identity management for unknown and time-varying number of targets in clutter

A novel multi-target identity management algorithm, called IM-GMPHD (Identity management with Gaussian mixture probability hypothesis density) algorithm, is proposed to simultaneously track an unknown and time-varying number of targets in clutter and manage the targets' identities effectively and efficiently over time. The targets' identities are managed by recursively computing the target identity beliefs using the target-measurement correlation. An augmented GMPHD filter is also developed in this paper to calculate the target-measurement correlation. In addition, the target identity beliefs can be updated to a lower uncertainty level whenever there is available target identity information. The proposed multi-target identity management algorithm is demonstrated through various illustrative simulation scenarios. (c) 2021 European Control Association. Published by Elsevier Ltd. All rights reserved.