A Public Video Dataset for Road Transportation Applications

Nicolas Saunier, Håkan Ardö, Jean-Philippe Jodoin, Aliaksei Laureshyn, Mikael Nilsson, Åse Svensson, Luis Miranda-Moreno, Guillaume-Alexandre Bilodeau, Karl Åström

Research output: Contribution to conferencePaper, not in proceedingpeer-review

421 Downloads (Pure)

Abstract

Video data and the tools for automated analysis have a great potential to be used in road traffic research, particularly road safety. In this project a video dataset is built and made public so that researchers can evaluate their algorithms on it. The dataset focuses on the traffic research applications (data from real research projects) and provides recordings of the traffic scenes, meta-data, camera calibration, ground truth, protocols for comparing algorithms and software tools and libraries for reading/presenting the data. To the authors’ knowledge, this public dataset is the first of its kind. With the proposed dataset, researchers get access to a large variety of recordings representing different traffic, weather and lighting conditions to evaluate and compare different tools and applications. As a consequence, discussions between computer vision and transportation researchers are expected to increase, contributing to more collaborations and better tools, more accurate and user-friendly, to obtain automatically rich traffic data from video.
Original languageEnglish
Number of pages17
Publication statusPublished - 2014
Event93th TRB Annual Meeting - Washington DC, United States
Duration: 2014 Jan 122014 Jan 16

Conference

Conference93th TRB Annual Meeting
Country/TerritoryUnited States
CityWashington DC
Period2014/01/122014/01/16

Subject classification (UKÄ)

  • Transport Systems and Logistics
  • Mathematics

Fingerprint

Dive into the research topics of 'A Public Video Dataset for Road Transportation Applications'. Together they form a unique fingerprint.

Cite this