Skip to content
Free planNo credit card requiredStart writing

article · Electronic Theses and Dissertations Repository (University of Pisa) · 2018

Reducing fragmentation in parameterized surfaces

MAGGIORDOMO, ANDREA

Cited by 0Open access

Abstract

This thesis faces the problem of improving the existing texture atlas of a polygonal model. We target textures generated by 3D reconstruction pipelines, which usually exhibit high fragmentation in the texture data, and try to reduce the number of of discontinuities without losing existing information. We do so by computing a new mesh parameterization that is less fragmented, has shorter boundaries and exhibits minimal distortion compared to the original texture map. Our approach relies on available parameterization algorithms, and adapts them in an original way to minimize deformation of the newly computed flattening with respect to the shape of the existing one. We report experimental results obtained by applying our method on a collection of publicly available models generated from 3D reconstruction software.

Cite this paper

APA 7

MAGGIORDOMO, A. (2018). Reducing fragmentation in parameterized surfaces. Electronic Theses and Dissertations Repository (University of Pisa).

MLA 9

MAGGIORDOMO, ANDREA. “Reducing fragmentation in parameterized surfaces.” Electronic Theses and Dissertations Repository (University of Pisa), 2018.

Chicago

MAGGIORDOMO, ANDREA. 2018. “Reducing fragmentation in parameterized surfaces.” Electronic Theses and Dissertations Repository (University of Pisa).

Harvard

MAGGIORDOMO, A. (2018) ‘Reducing fragmentation in parameterized surfaces’, Electronic Theses and Dissertations Repository (University of Pisa).

BibTeX
@article{maggiordomo2018reducing,
  title = {{Reducing fragmentation in parameterized surfaces}},
  author = {MAGGIORDOMO, ANDREA},
  journal = {Electronic Theses and Dissertations Repository (University of Pisa)},
  year = {2018}
}

Made from the index record. Check names, capitalisation and page numbers against the paper. Need another style? AI for Research has over 10,000.