A labeled random finite set spawning model

© 2017 IEEE. Previous labeled random finite set filter developments use a target motion model that only accounts for survival and birth. While such a model provides the means for a multi-target tracking filter such as the Generalized Labeled Multi-Bernoulli filter to capture target births and deaths...

Full description

Bibliographic Details
Main Authors: Bryant, D., Vo, Ba Tuong, Vo, Ba-Ngu, Jones, B.
Format: Conference Paper
Published: 2017
Online Access:http://purl.org/au-research/grants/arc/DP170104854
http://hdl.handle.net/20.500.11937/68138
Description
Summary:© 2017 IEEE. Previous labeled random finite set filter developments use a target motion model that only accounts for survival and birth. While such a model provides the means for a multi-target tracking filter such as the Generalized Labeled Multi-Bernoulli filter to capture target births and deaths in a wide variety of applications, it lacks the capability to capture the lineages of spawned target tracks. In this paper, we propose a labeled random finite set spawning model and derive the resulting multi-target prediction and filtering densities. This formulation enables the joint estimation of spawned object's state and and information regarding its lineage.