Publication Details

Feature extraction for efficient image and video segmentation

VOJVODA, J.; BERAN, V. Feature extraction for efficient image and video segmentation. In Proceedings - SCCG 2016: 32nd Spring Conference on Computer Graphics. Smolenice: Association for Computing Machinery, 2016. p. 75-80. ISBN: 978-1-4503-4436-4.
Czech title
Extrakce příznaků pro efektivní segmentaci obrazu a videa
Type
conference paper
Language
English
Authors
Vojvoda Jakub, Ing.
Beran Vítězslav, doc. Ing., Ph.D. (DCGM)
Keywords

Color/Texture segmentation, Motion segmentation, RGB-D/T segmentation

Abstract

The segmentation of sensory data of various domains is often crucial pre-processing step in many computer vision methods and applications. In this work, we propose a method that leverages the quantization of local features distributions for the depth and the temporal information. Three variants of the segmentation method is designed and evaluated reflecting various data domains: space (color and texture), temporal (motion) and depth domain. Each variant was tested on appropriate dataset showing the usability of designed method for applications like areal-image analysis, hand detection and moving-people detection. The pilot experiments shows the characteristics of the approach and computational costs of designed variants. 

Published
2016
Pages
75–80
Proceedings
Proceedings - SCCG 2016: 32nd Spring Conference on Computer Graphics
ISBN
978-1-4503-4436-4
Publisher
Association for Computing Machinery
Place
Smolenice
DOI
UT WoS
000403659700010
EID Scopus
BibTeX
@inproceedings{BUT130940,
  author="Jakub {Vojvoda} and Vítězslav {Beran}",
  title="Feature extraction for efficient image and video segmentation",
  booktitle="Proceedings - SCCG 2016: 32nd Spring Conference on Computer Graphics",
  year="2016",
  pages="75--80",
  publisher="Association for Computing Machinery",
  address="Smolenice",
  doi="10.1145/2948628.2948631",
  isbn="978-1-4503-4436-4",
  url="https://www.fit.vut.cz/research/publication/11086/"
}
Files
Back to top