Possibility Generalized Labeled Multi-Bernoulli Filter for Multitarget Tracking Under Epistemic Uncertainty

This article presents a flexible modeling framework for multitarget tracking based on the theory of outer probability measures. The notion of labeled uncertain finite set is introduced and utilized as the basis to derive a possibilistic analog of the $\delta$-generalized labeled multi-Bernoulli ($\delta$-GLMB) filter, in which the uncertainty in the multitarget system is represented by possibility functions instead of probability distributions. The proposed method inherits the capability of the standard probabilistic $\delta$-GLMB filter to yield joint state, number, and trajectory estimates of multiple appearing and disappearing targets. Beyond that, it is capable to account for epistemic uncertainty due to ignorance or partial knowledge regarding the multitarget system, e.g., the absence of complete information on dynamical model parameters (e.g., probability of detection, birth) and initial number and state of newborn targets. The features of the developed filter are demonstrated using two simulated scenarios.

Automated summary

This article presents a flexible modeling framework for multitarget tracking based on outer probability measures. It introduces the concept of labeled uncertain finite set and utilizes it to derive a possibilistic analog of the $\delta$-generalized labeled multi-Bernoulli filter. The proposed method can account for both uncertainty in the multitarget system and epistemic uncertainty due to partial knowledge or ignorance.