A particle dyeing approach for track continuity for the SMC-PHD filter

Tiancheng Li, Shudong Sun, Juan Manuel Corchado, Ming Fei Siyau

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

63 Scopus citations

Abstract

This paper proposes a novel particle labeling (termed as 'dyeing') method for track continuity for the sequential Monte Carlo (SMC) implementation of the probability hypothesis density (PHD) filter. The Multi-Expected a Posterior (MEAP) estimator is employed to extract estimates that is of high accuracy and fast computing speed. In the estimate extracting process, particles are dyed by the color of the closest observation (different observations have different color) that corresponds to an estimate or clutter. The estimates of two successive scans are then associated with respect to their dyeing color interaction on the particles. Unlike the general labeling method, not all particles will be labeled to an estimate/track in the dyeing process. No modification is required to make on the PHD equation due to dyeing/MEAP. The proposed estimate association method is able to handle track initialization, termination, maintenance including track splitting and merging, based on observations of successive scans.

Original languageEnglish
Title of host publicationFUSION 2014 - 17th International Conference on Information Fusion
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9788490123553
StatePublished - 3 Oct 2014
Event17th International Conference on Information Fusion, FUSION 2014 - Salamanca, Spain
Duration: 7 Jul 201410 Jul 2014

Publication series

NameFUSION 2014 - 17th International Conference on Information Fusion

Conference

Conference17th International Conference on Information Fusion, FUSION 2014
Country/TerritorySpain
CitySalamanca
Period7/07/1410/07/14

Keywords

  • labeling
  • Multi-target tracking
  • probability hypothesis density
  • sequential Monte Carlo
  • track continuity

Fingerprint

Dive into the research topics of 'A particle dyeing approach for track continuity for the SMC-PHD filter'. Together they form a unique fingerprint.

Cite this