Robust GMPHD filter with adaptive target birth

© 2014 IEEE. Recently, the Gaussian Mixture Probability Hypothesis Density (GMPHD) filter has been studied as a popular method for multi-target tracking in clutter background. As an extended version, robust GMPHD filter, also known as the ?-GMPHD filter, was subsequently proposed to handle unknown c...

Full description

Bibliographic Details
Main Authors: Ma, L., Wang, P., Xue, K., Kim, Du Yong
Format: Conference Paper
Published: 2014
Online Access:http://hdl.handle.net/20.500.11937/56402
_version_ 1848759866453131264
author Ma, L.
Wang, P.
Xue, K.
Kim, Du Yong
author_facet Ma, L.
Wang, P.
Xue, K.
Kim, Du Yong
author_sort Ma, L.
building Curtin Institutional Repository
collection Online Access
description © 2014 IEEE. Recently, the Gaussian Mixture Probability Hypothesis Density (GMPHD) filter has been studied as a popular method for multi-target tracking in clutter background. As an extended version, robust GMPHD filter, also known as the ?-GMPHD filter, was subsequently proposed to handle unknown clutter rate. In this letter, we use a Partially Uniform Birth (PUB) model for target births and present the ?-PUB-GMPHD filter, which accommodates tracking scenarios where targets can appear anywhere in the surveillance region and clutter rate is unknown. However, analysis of the ?-PUB-GMPHD filter demonstrates that clutter cardinality is underestimated due to the presence of PUB model and the cardinality bias of clutter is susceptible to spontaneous target birth rate. Hence, we propose adaptive birth rate technique for the ?-PUB-GMPHD filter to initialize new tracks quickly and simultaneously avoid severe cardinality bias. Simulations are present to verify the ?-PUB-GMPHD filter with proposed adaptive birth rate technique.
first_indexed 2025-11-14T10:06:41Z
format Conference Paper
id curtin-20.500.11937-56402
institution Curtin University Malaysia
institution_category Local University
last_indexed 2025-11-14T10:06:41Z
publishDate 2014
recordtype eprints
repository_type Digital Repository
spelling curtin-20.500.11937-564022017-09-13T16:11:34Z Robust GMPHD filter with adaptive target birth Ma, L. Wang, P. Xue, K. Kim, Du Yong © 2014 IEEE. Recently, the Gaussian Mixture Probability Hypothesis Density (GMPHD) filter has been studied as a popular method for multi-target tracking in clutter background. As an extended version, robust GMPHD filter, also known as the ?-GMPHD filter, was subsequently proposed to handle unknown clutter rate. In this letter, we use a Partially Uniform Birth (PUB) model for target births and present the ?-PUB-GMPHD filter, which accommodates tracking scenarios where targets can appear anywhere in the surveillance region and clutter rate is unknown. However, analysis of the ?-PUB-GMPHD filter demonstrates that clutter cardinality is underestimated due to the presence of PUB model and the cardinality bias of clutter is susceptible to spontaneous target birth rate. Hence, we propose adaptive birth rate technique for the ?-PUB-GMPHD filter to initialize new tracks quickly and simultaneously avoid severe cardinality bias. Simulations are present to verify the ?-PUB-GMPHD filter with proposed adaptive birth rate technique. 2014 Conference Paper http://hdl.handle.net/20.500.11937/56402 10.1109/ICCAIS.2014.7020556 restricted
spellingShingle Ma, L.
Wang, P.
Xue, K.
Kim, Du Yong
Robust GMPHD filter with adaptive target birth
title Robust GMPHD filter with adaptive target birth
title_full Robust GMPHD filter with adaptive target birth
title_fullStr Robust GMPHD filter with adaptive target birth
title_full_unstemmed Robust GMPHD filter with adaptive target birth
title_short Robust GMPHD filter with adaptive target birth
title_sort robust gmphd filter with adaptive target birth
url http://hdl.handle.net/20.500.11937/56402