The “Toyota Motor Europe (TME) Motorway Dataset” is composed by 28 clips for a total of approximately 27 minutes (30000+ frames) with vehicle annotation. Annotation was semi-automatically generated using laser-scanner data. Image sequences were selected from acquisition made in North Italian motorways in December 2011. This selection includes variable traffic situations, number of lanes, road curvature, and lighting, covering most of the conditions present in the complete acquisition.
The dataset comprises:
- Image acquisition: stereo, 20 Hz frequency , 1024x768 grayscale losslessly compressed images, 32° horizontal field of view, bayer coded color information (in OpenCV use CV_BayerGB2GRAY and CV_BayerGB2BGR color conversion codes; please note that left camera was rotated upside down, convert to color/grayscale BEFORE flipping the image). A checkboard calibration sequence is made available.
- Laser-scanner generated vehicle annotation and classification (car/truck).
- A software evaluation toolkit (C++ source code).
The data provided is timestamped, and includes extrinsic calibration.
The dataset has been divided in two sub-sets depending on lighting condition, named “daylight” (although with objects casting shadows on the road) and “sunset” (facing the sun or at dusk). For each clip, 5 seconds of preceding acquisition are provided, to allow the algorithm stabilizing before starting the actual performance measurement.
The data has been acquired in cooperation with VisLab (University of Parma, Italy), using the BRAiVE test vehicle.
Release
This Computer Vision dataset is publicly released to promote comparison. The images from the right camera (on which results should be computed) and left camera for the full dataset are available for download at this link. For privacy reasons, they are stored in password protected archives. The use of the data is free in the context of scientific research (note: use for research in commercial domain is equally allowed).
To obtain the password to the image archives preferably contact:
<first author name>.<first author surname>@gmail.com
<first author name>.<first author surname>@toyota-europe.com
or, alternatively, one of the other authors (see paper below). This will also allow us to inform you about updates. By requesting the password you acknowledge the limitation of use for scientific purpose.
The laser-scanner generated annotation, together with the C++ source code of the evaluation software, is available for download at this link. Check included instructions. For explanation about the mechanisms implemented see paper, paragraph IV. Click here to navigate the source code.
The rest of the dataset (egomotion, image timestamps) will be made available upon request.
To obtain the password to the image archives preferably contact:
<first author name>.<first author surname>@gmail.com
<first author name>.<first author surname>@toyota-europe.com
or, alternatively, one of the other authors (see paper below). This will also allow us to inform you about updates. By requesting the password you acknowledge the limitation of use for scientific purpose.
The laser-scanner generated annotation, together with the C++ source code of the evaluation software, is available for download at this link. Check included instructions. For explanation about the mechanisms implemented see paper, paragraph IV. Click here to navigate the source code.
The rest of the dataset (egomotion, image timestamps) will be made available upon request.
Publication
In case you use this dataset in an academic work, please cite the following paper:
@InProceedings{ TMEMotorwayDataset,
author = "Claudio Caraffi and Tomas Vojir and Jura Trefny and Jan Sochman and Jiri Matas",
title = "{A System for Real-time Detection and Tracking of Vehicles from a Single Car-mounted Camera}",
booktitle = "ITS Conference",
year = "2012",
pages = "975--982",
month = "Sep."
}
Download links: paper, bibtex file, presentation
author = "Claudio Caraffi and Tomas Vojir and Jura Trefny and Jan Sochman and Jiri Matas",
title = "{A System for Real-time Detection and Tracking of Vehicles from a Single Car-mounted Camera}",
booktitle = "ITS Conference",
year = "2012",
pages = "975--982",
month = "Sep."
}
Video
Persistent vehicle detection and tracking from a single camera in motorway scenario,
in different traffic and varying lighting condition.
Results are compared with a "Ground Truth" automatically generated from laserscanner data.
See Readme file for full explanation.
The video can be download from this link.
The video can be download from this link.
This material is
presented to ensure timely dissemination of scholarly and technical work.
Copyright and all rights therin are retained by authors or by other copyright
holders. All persons copying this information are expected to adhere to the
terms and constraints invoked by each author's copyright. In most cases,
these works may not be reposted without the explicit permission of the
copyright holder. Personal use of this material is permitted. However,
permission to reprint/republish this material for advertising or promotional
purposes or for creating new collective works for resale or redistribution to
servers or lists, or to reuse any copyrighted component of the works
published in IEEE publications in other works must be obtained from the IEEE.