Vision-Guided Robot Tracking Algorithm Based on Characteristic Description of Maneuvering Target

Yuan Zhang, Peng Wang, Xin Li, Shaochen Kang, Jinglei Xin, Wenhao Jiang

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

1 Scopus citations

Abstract

The change of appearance characteristic is caused by changing viewpoint in the maneuvering target tracking. The traditional Camshift tracking algorithm based on a single color histogram model is not robust to this variation. In this paper, image moment invariant is proposed to describe maneuvering target to enhance model description, through its recursive algorithm to optimize the target model and refresh the weight value. The experiment result demonstrates that, comparing with the single fixed model and the self-adaption single model, the new algorithm can adapt for the appearance variation of the target rapidly and with small calculated amount.

Original languageEnglish
Title of host publicationIntelligent Computing and Information Science - International Conference, ICICIS 2011, Proceedings
EditorsRan Chen
Pages731-736
Number of pages6
EditionPART 1
DOIs
StatePublished - 2011
Externally publishedYes
Event2011 International Conference on Intelligent Computing and Information Science, ICICIS 2011 - Chongqing, China
Duration: 8 Jan 20119 Jan 2011

Publication series

NameCommunications in Computer and Information Science
NumberPART 1
Volume134
ISSN (Print)1865-0929

Conference

Conference2011 International Conference on Intelligent Computing and Information Science, ICICIS 2011
Country/TerritoryChina
CityChongqing
Period8/01/119/01/11

Keywords

  • Camshift
  • Target tracking
  • Vision-guided robot
  • invariant moment

Fingerprint

Dive into the research topics of 'Vision-Guided Robot Tracking Algorithm Based on Characteristic Description of Maneuvering Target'. Together they form a unique fingerprint.

Cite this