Publication detail
Automatic Camera Calibration for Traffic Understanding
JURÁNKOVÁ, M. SOCHOR, J. HEROUT, A.
Original Title
Automatic Camera Calibration for Traffic Understanding
English Title
Automatic Camera Calibration for Traffic Understanding
Type
conference paper
Language
en
Original 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.
English 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.
Keywords
camera calibration, speed measurement, vanishing point detection, surveillance camera, traffic analysis
RIV year
2014
Released
08.07.2014
Publisher
The British Machine Vision Association and Society for Pattern Recognition
Location
Nottingham
ISBN
1-901725-52-9
Book
Proceedings of BMVC 2014
Edition
NEUVEDEN
Edition number
NEUVEDEN
Pages from
1
Pages to
10
Pages count
9
URL
Documents
BibTex
@inproceedings{BUT111627,
author="Markéta {Juránková} and Jakub {Sochor} and Adam {Herout}",
title="Automatic Camera Calibration for Traffic Understanding",
annote="
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.",
address="The British Machine Vision Association and Society for Pattern Recognition",
booktitle="Proceedings of BMVC 2014",
chapter="111627",
edition="NEUVEDEN",
howpublished="online",
institution="The British Machine Vision Association and Society for Pattern Recognition",
year="2014",
month="july",
pages="1--10",
publisher="The British Machine Vision Association and Society for Pattern Recognition",
type="conference paper"
}