Motion estimation for 3D rigid object shapes based on 3D-2D color-consistency in multi-view

Xiuxiu Li, Jiangbin Zheng, Huan Xu

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

Abstract

A motion estimation algorithm for 3D rigid object in multi-view is proposed in this paper. In the algorithm, firstly, calculate the color difference between the shape of the 3D object and the silhouettes from multi-view to get the 3D-2D color-consistency; secondly, define an object function about the 3D- 2D color-consistency; lastly, minimize the object function with Levenberg- Marquard algorithm to obtain the 3D motion parameters. It is obvious that the algorithm could not be affected by the distribution and positioning accuracy of features owing to the usage of the whole color information. Some experiments show that the algorithm for the 3D shape motion estimation in multi-view videos is effective.

Original languageEnglish
Title of host publicationIntelligent Science and Intelligent Data Engineering - Second Sino-Foreign-Interchange Workshop, IScIDE 2011, Revised Selected Papers
Pages450-457
Number of pages8
DOIs
StatePublished - 2012
Event2nd Sino-Foreign-Interchange Workshop on Intelligent Science and Intelligent Data Engineering, IScIDE 2011 - Xi'an, China
Duration: 23 Oct 201125 Oct 2011

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7202 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd Sino-Foreign-Interchange Workshop on Intelligent Science and Intelligent Data Engineering, IScIDE 2011
Country/TerritoryChina
CityXi'an
Period23/10/1125/10/11

Keywords

  • 3D shapes motion estimation
  • color-consistency
  • Multi-view

Fingerprint

Dive into the research topics of 'Motion estimation for 3D rigid object shapes based on 3D-2D color-consistency in multi-view'. Together they form a unique fingerprint.

Cite this