Publication Details

GeoPose3K: Mountain Landscape Dataset for Camera Pose Estimation in Outdoor Environments

BREJCHA Jan and ČADÍK Martin. GeoPose3K: Mountain Landscape Dataset for Camera Pose Estimation in Outdoor Environments. Image and Vision Computing, vol. 2017, no. 1, pp. 1-41. ISSN 0262-8856. Available from: https://www.sciencedirect.com/science/article/abs/pii/S0262885617300963?via%3Dihub
Czech title
GeoPose3K: Dataset horské krajiny pro odhad pozice a orientace kamery v přírodním prostředí
Type
journal article
Language
english
Authors
Brejcha Jan, Ing., Ph.D. (DCGM FIT BUT)
Čadík Martin, doc. Ing., Ph.D. (DCGM FIT BUT)
URL
Keywords

camera pose estimation, visual geo-localization,camera orientation estimation,image-to-model registration, digital terrain models, semantic segmentation 

Abstract

We introduce a new dataset called GeoPose3K which contains over three thousand precise camera poses of mountain landscape images. In addition to camera location and orientation, we provide data for training and evaluation of computer vision methods and applications in context of outdoor scenes; synthetic depth maps, normal maps, illumination simulation and semantic labels. To illustrate properties of the dataset, we compare results achieved by state-of-the-art visual geo-localization method on GeoPose3K with results achieved on an existing dataset for visual geo-localization. To foster research of computer vision algorithms for outdoor environments, we propose several novel future use-cases of our new GeoPose3K dataset. 

Published
2017
Pages
1-41
Journal
Image and Vision Computing, vol. 2017, no. 1, ISSN 0262-8856
Publisher
Elsevier Science
DOI
10.1016/j.imavis.2017.05.009
UT WoS
000413060000001
EID Scopus
2-s2.0-85028660327
BibTeX 
@ARTICLE{FITPUB11463,
   author = "Jan Brejcha and Martin \v{C}ad\'{i}k",
   title = "GeoPose3K: Mountain Landscape Dataset for Camera Pose Estimation in Outdoor Environments",
   pages = "1--41",
   journal = "Image and Vision Computing",
   volume = 2017,
   number = 1,
   year = 2017,
   ISSN = "0262-8856",
   doi = "10.1016/j.imavis.2017.05.009",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/11463"
}
Files
Back to top