Skip to main navigation Skip to search Skip to main content

Group tracking based on hypothesis management aided MCMC-PF algorithm

  • Beihang University

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

Abstract

Group tracking based on Markov Chain Monte Carlo particle filter (MCMC-PF) algorithm has high accuracy of individual target tracking and group tracking, but the main problem of MCMC-PF algorithm is calculation burden, that's because it considers all kinds of hypotheses between state vector and measurement vector in the observation model. Therefore, in this paper, we originally use the results of data association as a prior to do the hypotheses management, which will reduce the variety of feasible hypotheses. We also raise a new method for the target state proposal based on group information and Kalman framework. In the end, simulation results demonstrate that the new algorithms can track the individual target and infer the correct group structure with less time consumption than previous work.

Original languageEnglish
Title of host publication2016 CIE International Conference on Radar, RADAR 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509048281
DOIs
StatePublished - 4 Oct 2017
Event2016 CIE International Conference on Radar, RADAR 2016 - Guangzhou, China
Duration: 10 Oct 201613 Oct 2016

Publication series

Name2016 CIE International Conference on Radar, RADAR 2016

Conference

Conference2016 CIE International Conference on Radar, RADAR 2016
Country/TerritoryChina
CityGuangzhou
Period10/10/1613/10/16

Keywords

  • Group tracking
  • Hypothesis management
  • Kalman
  • MCMC-PF algorithm

Fingerprint

Dive into the research topics of 'Group tracking based on hypothesis management aided MCMC-PF algorithm'. Together they form a unique fingerprint.

Cite this