Diminished reality via multiple hand-held cameras

Songkran Jarusirisawad, Hideo Saito

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

This paper proposes a novel method for calibrating multiple hand-held cameras target for Diminished Reality application. Our method does not require any special markers or information about camera parameters. Projective Grid Space (PGS) which is 3D space defined by epipolar geometry of two basis camera is used for dynamic cameras calibration. Geometrical relations among cameras in PGS are obtained from 2D-2D corresponding points between views. We utilize Scale Invariant Feature Transform (SIFT) for finding corresponding points in natural scene for registering cameras to PGS. Moving object is segmented via graph cut optimization. Finally, the reconstructed visual hull is used to synthesize free viewpoint video in which unwanted or occluding object is deliberately removed. In the experimental results, free viewpoint video without unwanted object which is captured by hand-held cameras is successfully synthesized using the proposed method.

Original languageEnglish
Title of host publication2007 1st ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC
Pages251-258
Number of pages8
DOIs
Publication statusPublished - 2007 Dec 1
Event2007 First ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC-07 - Vienna, Austria
Duration: 2007 Sept 252007 Sept 28

Publication series

Name2007 1st ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC

Other

Other2007 First ACM/IEEE International Conference on Distributed Smart Cameras, ICDSC-07
Country/TerritoryAustria
CityVienna
Period07/9/2507/9/28

Keywords

  • 3D reconstruction
  • Cameras calibration
  • Diminished reality
  • Projective grid space

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'Diminished reality via multiple hand-held cameras'. Together they form a unique fingerprint.

Cite this