Dense 3D reconstruction from images by normal aided matching

Zoltán Megyesi, Géza Kós, D. Chetverikov

Research output: Contribution to journalArticle

22 Citations (Scopus)

Abstract

3D models play an increased role in today's computer applications. As a result, there is a need for flexible and easy to use measuring devices that produce 3D models of real world objects. 3D scene reconstruction is a quickly evolving field of computer vision, which aims at creating 3D models from images of a scene. Although many problems of the reconstruction process have been solved, the use of photographs as an information source involves some practical difficulties. Therefore, accurate and dense 3D reconstruction remains a challenging task. We discuss dense matching of surfaces in the case when the images are taken from a wide baseline camera setup. Some recent studies use a region-growing based dense matching framework, and improve accuracy through estimating the apparent distortion by local affine transformations. In this paper we present a way of using pre-calculated calibration data to improve precision. We demonstrate that the new method produces a more accurate model.

Original languageEnglish
Pages (from-to)3-28
Number of pages26
JournalMachine Graphics and Vision
Volume15
Issue number1
Publication statusPublished - 2006

Fingerprint

Computer applications
Computer vision
Cameras
Calibration

Keywords

  • 3D scene reconstruction
  • Dense matching
  • Image based reconstruction
  • Stereo
  • Wide-baseline

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Computer Vision and Pattern Recognition

Cite this

Dense 3D reconstruction from images by normal aided matching. / Megyesi, Zoltán; Kós, Géza; Chetverikov, D.

In: Machine Graphics and Vision, Vol. 15, No. 1, 2006, p. 3-28.

Research output: Contribution to journalArticle

Megyesi, Zoltán ; Kós, Géza ; Chetverikov, D. / Dense 3D reconstruction from images by normal aided matching. In: Machine Graphics and Vision. 2006 ; Vol. 15, No. 1. pp. 3-28.
@article{510b6f45fcd54ba693ebad88015cc69f,
title = "Dense 3D reconstruction from images by normal aided matching",
abstract = "3D models play an increased role in today's computer applications. As a result, there is a need for flexible and easy to use measuring devices that produce 3D models of real world objects. 3D scene reconstruction is a quickly evolving field of computer vision, which aims at creating 3D models from images of a scene. Although many problems of the reconstruction process have been solved, the use of photographs as an information source involves some practical difficulties. Therefore, accurate and dense 3D reconstruction remains a challenging task. We discuss dense matching of surfaces in the case when the images are taken from a wide baseline camera setup. Some recent studies use a region-growing based dense matching framework, and improve accuracy through estimating the apparent distortion by local affine transformations. In this paper we present a way of using pre-calculated calibration data to improve precision. We demonstrate that the new method produces a more accurate model.",
keywords = "3D scene reconstruction, Dense matching, Image based reconstruction, Stereo, Wide-baseline",
author = "Zolt{\'a}n Megyesi and G{\'e}za K{\'o}s and D. Chetverikov",
year = "2006",
language = "English",
volume = "15",
pages = "3--28",
journal = "Machine Graphics and Vision",
issn = "1230-0535",
publisher = "Polska Akademia Nauk",
number = "1",

}

TY - JOUR

T1 - Dense 3D reconstruction from images by normal aided matching

AU - Megyesi, Zoltán

AU - Kós, Géza

AU - Chetverikov, D.

PY - 2006

Y1 - 2006

N2 - 3D models play an increased role in today's computer applications. As a result, there is a need for flexible and easy to use measuring devices that produce 3D models of real world objects. 3D scene reconstruction is a quickly evolving field of computer vision, which aims at creating 3D models from images of a scene. Although many problems of the reconstruction process have been solved, the use of photographs as an information source involves some practical difficulties. Therefore, accurate and dense 3D reconstruction remains a challenging task. We discuss dense matching of surfaces in the case when the images are taken from a wide baseline camera setup. Some recent studies use a region-growing based dense matching framework, and improve accuracy through estimating the apparent distortion by local affine transformations. In this paper we present a way of using pre-calculated calibration data to improve precision. We demonstrate that the new method produces a more accurate model.

AB - 3D models play an increased role in today's computer applications. As a result, there is a need for flexible and easy to use measuring devices that produce 3D models of real world objects. 3D scene reconstruction is a quickly evolving field of computer vision, which aims at creating 3D models from images of a scene. Although many problems of the reconstruction process have been solved, the use of photographs as an information source involves some practical difficulties. Therefore, accurate and dense 3D reconstruction remains a challenging task. We discuss dense matching of surfaces in the case when the images are taken from a wide baseline camera setup. Some recent studies use a region-growing based dense matching framework, and improve accuracy through estimating the apparent distortion by local affine transformations. In this paper we present a way of using pre-calculated calibration data to improve precision. We demonstrate that the new method produces a more accurate model.

KW - 3D scene reconstruction

KW - Dense matching

KW - Image based reconstruction

KW - Stereo

KW - Wide-baseline

UR - http://www.scopus.com/inward/record.url?scp=33747651067&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33747651067&partnerID=8YFLogxK

M3 - Article

VL - 15

SP - 3

EP - 28

JO - Machine Graphics and Vision

JF - Machine Graphics and Vision

SN - 1230-0535

IS - 1

ER -