Histogram-Probabilistic Multi-Hypothesis Tracking with Integrated Target Existence
Abstract
The histogram-probabilistic multi-hypothesis tracker (H-PMHT) is a parametric approach to solving the multi-target track-before-detect (TBD) problem, using expectation maximisation (EM). A key limitation of this method is the assumption of a known and constant number of targets. In this paper, we propose the integrated existence Poisson histogram probabilistic multi-hypothesis tracker (IE-PHPMHT), for TBD of multiple targets. It extends the H-PMHT framework by adding a probability of existence to each potential target. For the derivation, we utilise a Poisson point process (PPP) measurement model and Bernoulli targets, allowing for a multi-Bernoulli birth process and an unknown, time-varying number of targets. Hence, integrated track management is achieved through the discrimination of track quality assessments based on existence probabilities. The algorithm is evaluated in a simulation study of two scenarios and is compared with several other algorithms, demonstrating its performance.
Keywords
Cite
@article{arxiv.2504.20526,
title = {Histogram-Probabilistic Multi-Hypothesis Tracking with Integrated Target Existence},
author = {Lukas Herrmann and Ángel F. García-Fernández and Edmund F. Brekke},
journal= {arXiv preprint arXiv:2504.20526},
year = {2025}
}
Comments
Submitted for possible publication in IEEE Transactions on Aerospace and Electronic Systems (TAES)