Publication Details

Automatic Camera Calibration for Traffic Understanding

DUBSKÁ Markéta, SOCHOR Jakub and HEROUT Adam. Automatic Camera Calibration for Traffic Understanding. In: Proceedings of BMVC 2014. Nottingham: The British Machine Vision Association and Society for Pattern Recognition, 2014, pp. 1-10. ISBN 1-901725-52-9.
Czech title
Automatická kalibrace kamery pro zpracování silničních dat
Type
conference paper
Language
english
Authors
Dubská Markéta, Ing., Ph.D. (DCGM FIT BUT)
Sochor Jakub, Ing. (DCGM FIT BUT)
Herout Adam, prof. Ing., Ph.D. (DCGM FIT BUT)
URL
Keywords

camera calibration, speed measurement, vanishing point detection, surveillance camera, traffic analysis

Abstract


We propose a method for fully automatic calibration of traffic surveillance cameras. This method allows for calibration of the camera - including scale - without any user input, only from several minutes of input surveillance video. The targeted applications include speed measurement, measurement of vehicle dimensions, vehicle classification, etc. 
The achieved mean accuracy of speed and distance measurement is below 2%. Our efficient C++ implementation runs in real time on a lowend processor (Core i3) with a safe margin even for full-HD videos.

Published
2014
Pages
1-10
Proceedings
Proceedings of BMVC 2014
Conference
British Machine Vision Conference (BMVC) 2014, Nottingham, GB
ISBN
1-901725-52-9
Publisher
The British Machine Vision Association and Society for Pattern Recognition
Place
Nottingham, GB
EID Scopus
2-s2.0-84919772000
BibTeX 
@INPROCEEDINGS{FITPUB10682,
   author = "Mark\'{e}ta Dubsk\'{a} and Jakub Sochor and Adam Herout",
   title = "Automatic Camera Calibration for Traffic Understanding",
   pages = "1--10",
   booktitle = "Proceedings of BMVC 2014",
   year = 2014,
   location = "Nottingham, GB",
   publisher = "The British Machine Vision Association and Society for Pattern Recognition",
   ISBN = "1-901725-52-9",
   language = "english",
   url = "https://www.fit.vut.cz/research/publication/10682"
}
Back to top