Skip to main navigation Skip to search Skip to main content

Simplification for texture mapping models with mesh segmentation

  • Lili Wang*
  • , Zhiqiang Ma
  • , Bing Xue
  • , Zhe Shen
  • *Corresponding author for this work
  • Beihang University

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

Abstract

To reduce the loss of visual details in simplification for textured meshes, this paper presents a simplification method based on mesh segmentation. Firstly, the model is partitioned into some sub-meshes according to their properties of appearance, such as texture, normal, etc. Thus, the correlation between sub-meshes can be decreased. Then, we calculate the error metric for the edge collapse in each sub-mesh. Lastly, we merge the simplified sub-meshes and fix the gap among them to regain an integrated mesh. Because the properties of appearance are used to separate the high-frequency details from low-frequency ones, our approach decreases the loss of the detailed. Therefore, we can not only keep the over whole contours of the model, but also maintain better geometric and appearance details. The results of experiments demonstrate our mesh segmentation based method is effective.

Original languageEnglish
Title of host publication2010 16th International Conference on Virtual Systems and Multimedia, VSMM 2010
Pages197-203
Number of pages7
DOIs
StatePublished - 2010
Event2010 16th International Conference on Virtual Systems and Multimedia, VSMM 2010 - Seoul, Korea, Republic of
Duration: 20 Oct 201023 Oct 2010

Publication series

Name2010 16th International Conference on Virtual Systems and Multimedia, VSMM 2010

Conference

Conference2010 16th International Conference on Virtual Systems and Multimedia, VSMM 2010
Country/TerritoryKorea, Republic of
CitySeoul
Period20/10/1023/10/10

Keywords

  • Details preservation
  • Mesh segmentation
  • Mesh simplification

Fingerprint

Dive into the research topics of 'Simplification for texture mapping models with mesh segmentation'. Together they form a unique fingerprint.

Cite this