A Sequential Approach to Capture Fine Geometric Details from Images

Download
  1. (PDF, 1 MB)
AuthorSearch for:
TypeArticle
ConferenceProceedings of ISPRS Commission V Symposium - Image Engineering and Vision Metrology, September 25-27, 2006.
VolumeVolume XXXVI, Part 5
Subjecthigh resolution; matching; modelling; photo-realism; rendering; texture; three-dimensional; visualization
AbstractWe present a sequential multi-stage segment-based modelling from images procedure designed to model the fine geometric details with high-resolution meshes. Results from one stage guide the next stage. The objective is to increase the geometric details at the modelling phase in an automatic and reliable manner to achieve photo-realistic visualisation for a variety of applications. In the first stage of this approach, basic shapes are determined with an interactive approach while in the subsequent stages fine details are added automatically, on a segment by segment basis, using dense stereo matching and shape from shading. The two techniques are used in a complementary manner and selecting which one to use is based on preset logical rules. Since we only need to compute the difference between the detailed model and the basic model produced in the first stage, the approach works in a more robust and reliable manner as compared to computing absolute coordinates. The approach has been extensively tested on hundreds of actual data and the results proved that it is effective, stable, and can model many types of surface details.
Publication date
LanguageEnglish
AffiliationNRC Institute for Information Technology; National Research Council Canada
Peer reviewedNo
NRC number48735
NPARC number5764790
Export citationExport as RIS
Report a correctionReport a correction
Record identifierfd6c8349-d3f3-47f0-a222-14ea835d952a
Record created2009-03-29
Record modified2016-05-09
Bookmark and share
  • Share this page with Facebook (Opens in a new window)
  • Share this page with Twitter (Opens in a new window)
  • Share this page with Google+ (Opens in a new window)
  • Share this page with Delicious (Opens in a new window)