A research of targets tracking and positioning algorithm based on multi-feature fusion

  • Pengfeng Chen
  • , Long Zhao*
  • *Corresponding author for this work

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

Abstract

In order to improve the accuracy and robustness of targets tracking and positioning, this paper proposes a particle filter algorithm based on multi-feature fusion. According to the diversity of character information, a multi-feature fusion strategy based on color, texture, and edge character has been developed, which can realize the comprehensive utilization of various visual features information. A feature criterion function has been addressed to the weighted strategy of multi-feature fusion, which can adjust adaptively the weight of feature and enhance the reliability of target tracking. Combined with binocular stereo vision technology, this algorithm can locate the target by calculating the geometrical relationship between the correspondence pixel points and spatial points. The test results show that the algorithm can realize the target tracking and positioning more accurately.

Original languageEnglish
Title of host publicationChina Satellite Navigation Conference, CSNC 2016, Proceedings
EditorsJingnan Liu, Shiwei Fan, Jiadong Sun, Feixue Wang
PublisherSpringer Verlag
Pages333-343
Number of pages11
ISBN (Print)9789811009334
DOIs
StatePublished - 2016
Event7th China Satellite Navigation Conference, CSNC 2016 - Changsha, China
Duration: 18 May 201620 May 2016

Publication series

NameLecture Notes in Electrical Engineering
Volume388
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

Conference7th China Satellite Navigation Conference, CSNC 2016
Country/TerritoryChina
CityChangsha
Period18/05/1620/05/16

Keywords

  • Adaptive weights
  • Binocular stereo vision
  • Multi-feature fusion
  • Particle filter
  • Target positioning
  • Target tracking

Fingerprint

Dive into the research topics of 'A research of targets tracking and positioning algorithm based on multi-feature fusion'. Together they form a unique fingerprint.

Cite this