3D 'micro'-geometry modeling from image cues

Yu Wang*, Charlie C.L. Wang, Matthew M.F. Yuen

*Corresponding author for this work

Research output: Chapter in Book/Conference Proceeding/ReportConference Paper published in a bookpeer-review

Abstract

This paper presents an approach to construct 3D microstructure of a surface from a single image. By integrating our technique with traditional CAD tools, the surface modeling results have clear 'micro' surface structures. Firstly, a multi-pyramid image segmentation algorithm is applied on the single image to extract the objective region. Our segmentation algorithm is an energy-based model, and it is accelerated by multi-pyramid. In the following, we use a modified shape from shading (SFS) model to construct the 3D surface described in the extracted image region. Our method is a global SFS approach, which is based on a new energy function that represents the shading difference between the reconstructed image (obtained from the constructed 3D object) and the input image.

Original languageEnglish
Title of host publicationProceedings of the ASME Design Engineering Technical Conference
PublisherAmerican Society of Mechanical Engineers
Pages589-596
Number of pages8
ISBN (Electronic)0791836215
DOIs
Publication statusPublished - 2002
Event22nd Computers and Information in Engineering Conference - Montreal, Que., Canada
Duration: 29 Sept 20022 Oct 2002

Publication series

NameProceedings of the ASME Design Engineering Technical Conference
Volume1

Conference

Conference22nd Computers and Information in Engineering Conference
Country/TerritoryCanada
CityMontreal, Que.
Period29/09/022/10/02

Keywords

  • Geometric modeling
  • Microstructure
  • Region extraction
  • Single image
  • Surface construction

Fingerprint

Dive into the research topics of '3D 'micro'-geometry modeling from image cues'. Together they form a unique fingerprint.

Cite this