Region-based epipolar and planar geometry estimation in low-textured environment

Vincent Gaudillière 1 Gilles Simon 1 Marie-Odile Berger 1
1 MAGRIT - Visual Augmentation of Complex Environments
Inria Nancy - Grand Est, LORIA - ALGO - Department of Algorithms, Computation, Image and Geometry
Abstract : Given two views of the same scene, usual correspondence geometry estimation techniques classically exploit the well-established effectiveness of keypoint descriptors. However, such features have a hard time in poorly textured man-made environments, possibly containing repetitive patterns and/or specularities, such as industrial places. In that paper, we propose a novel method for two-view epipolar and planar geometry estimation that first aims at detecting and matching physical vertical planes frequently present in these environments, before estimating corresponding homographies. Inferred local correspondences are finally used to improve fundamental matrix estimation. The gain in precision is demonstrated on industrial and urban environments.
Complete list of metadatas

Cited literature [23 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01802596
Contributor : Vincent Gaudilliere <>
Submitted on : Tuesday, May 29, 2018 - 2:52:00 PM
Last modification on : Monday, January 6, 2020 - 1:29:21 PM
Long-term archiving on: Thursday, August 30, 2018 - 5:58:53 PM

File

ICIP2018_Gaudilliere.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01802596, version 1

Collections

Citation

Vincent Gaudillière, Gilles Simon, Marie-Odile Berger. Region-based epipolar and planar geometry estimation in low-textured environment. ICIP 2018 - 25th IEEE International Conference on Image Processing, Oct 2018, Athens, Greece. pp.898-902. ⟨hal-01802596⟩

Share

Metrics

Record views

321

Files downloads

388