Detail publikace

Camera Orientation Estimation in Natural Scenes Using Semantic Cues

BREJCHA, J. ČADÍK, M.

Originální název

Camera Orientation Estimation in Natural Scenes Using Semantic Cues

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

angličtina

Originální abstrakt

Camera orientation estimation in natural scenes has recently been approached by several methods, which rely mainly on matching a single modality - edges or horizon lines with 3D digital elevation models. In contrast to previous works, our new image to model matching scheme is based on a fusion of multiple modalities and is designed to be naturally extensible with different cues. In this paper, we use semantic segments and edges. To our knowledge, we are the first to consider using semantic segments jointly with edges for alignment with digital elevation model. We show that high-level features, such as semantic segments, complement the low-level edge information and together help to estimate the camera orientation more robustly compared to methods relying solely on edges or horizon lines. In a series of experiments, we show that segment boundaries tend to be imprecise and important information for matching is encoded in the segment area and a coarse shape. Intuitively, semantic segments encode low frequency information as opposed to edges, which encode high frequencies. Our experiments exhibit that semantic segments and edges are complementary, improving camera orientation estimation reliability when used together. We demonstrate that our method combining semantic and edge features is able to reach state-of-the-art performance on three datasets.

Klíčová slova

camera orientation estimation, camera calibration, semantic segmentation, digital elevation model of a terrain, OpenStreetMap, geo-localization, computer vision, computer graphics

Autoři

BREJCHA, J.; ČADÍK, M.

Vydáno

21. 9. 2018

Nakladatel

IEEE Computer Society

Místo

Verona

ISBN

978-1-5386-2610-8

Kniha

2018 International Conference on 3D Vision

Strany od

208

Strany do

217

Strany počet

10

URL

https://www.fit.vut.cz/research/publication/11829/

BibTex

@inproceedings{BUT155091,
  author="Jan {Brejcha} and Martin {Čadík}",
  title="Camera Orientation Estimation in Natural Scenes Using Semantic Cues",
  booktitle="2018 International Conference on 3D Vision",
  year="2018",
  pages="208--217",
  publisher="IEEE Computer Society",
  address="Verona",
  doi="10.1109/3DV.2018.00033",
  isbn="978-1-5386-2610-8",
  url="https://www.fit.vut.cz/research/publication/11829/"
}