Possibility Generalized Labeled Multi-Bernoulli Filter for Multi-Target Tracking Under Epistemic Uncertainty
Han Cai, Jérémie Houssineau, Brandon A. Jones, Moriba Jah, Jingrui Zhang
Abstract
This paper presents a flexible modeling framework for multi-target tracking based on the theory of Outer Probability Measures (OPMs). The notion of labeled uncertain finite set is introduced and utilized as the basis to derive a possibilistic analog of the <inline-formula><tex-math notation="LaTeX">$\delta$</tex-math></inline-formula>-Generalized Labeled Multi-Bernoulli (<inline-formula><tex-math notation="LaTeX">$\delta$</tex-math></inline-formula>-GLMB) filter, in which the uncertainty in the multi-target system is represented by possibility functions instead of probability distributions. The proposed method inherits the capability of the standard probabilistic <inline-formula><tex-math notation="LaTeX">$\delta$</tex-math></inline-formula>-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 multi-target 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.