The Marginalized $\delta$-GLMB Filter

TL;DR

This paper introduces an efficient approximation to the $ ext{delta}$-GLMB filter that maintains key statistical properties for multi-target tracking, enabling scalable multi-sensor applications with improved computational efficiency.

Contribution

The paper proposes a novel approximation to the $ ext{delta}$-GLMB filter that preserves the PHD and cardinality distributions, addressing computational growth issues.

Findings

The approximation effectively maintains the PHD and cardinality distributions.

Simulation results demonstrate improved efficiency and accuracy in multi-sensor tracking.

The method scales better with the number of targets and sensors.

Abstract

The multi-target Bayes filter proposed by Mahler is a principled solution to recursive Bayesian tracking based on RFS or FISST. The $\delta$-GLMB filter is an exact closed form solution to the multi-target Bayes recursion which yields joint state and label or trajectory estimates in the presence of clutter, missed detections and association uncertainty. Due to presence of explicit data associations in the $\delta$-GLMB filter, the number of components in the posterior grows without bound in time. In this work we propose an efficient approximation to the $\delta$-GLMB filter which preserves both the PHD and cardinality distribution of the labeled posterior. This approximation also facilitates efficient multi-sensor tracking with detection-based measurements. Simulation results are presented to verify the proposed approach.