IMM-Based PMBM Filter for Maneuvering Extended Target Tracking
Jiayi Gu, Jiaqi Shi, Kun Shi, Chaoqun Yang, Xianghui Cao · 2024
In the real-world scenarios of maneuvering extended target tracking (METT), targets always exhibit complex motion patterns, making it challenging for traditional tracking methods with single motion model to accurately capture their trajectories. To address this challenge, this paper proposes an interacting multiple model (IMM) based Poisson multi-Bernoulli mixture (PMBM) filter to track maneuvering extended targets. In this paper, we first introduce the basic model of the maneuvering extended targets. Then, we present the proposed IMM-based PMBM filter in detail. Finally, simulation experiments are conducted, which demonstrate the superior performance of the proposed filter in the term of tracking multiple maneuvering extended targets.