Skip to content
Branch: master
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

ELAS Database

This dataset was recorded using a GoPRO HERO 3 camera in Brazil in different days at 29.97 frames per second. It comprises 22 scenes (more than 15,000 frames) containing all sorts of challenges we expect to encounter in the real world: highways, urban road, traffic, intersections, lane changes, writings on the road, shadows, wobbly capture, different weather, slightly different camera positions, faded lane markings, multiple LMT transitions, different vehicle speed, etc. Scenes were recorded in three cities: Vitória, Vila Velha, and Guarapari on the state of Espírito Santo, Brazil. Each frame has 640x480 pixels. This dataset was manually annotated for relevant events to the research community, such as: lane estimation, change, and centering; road markings; intersections; LMTs; crosswalks; and adjacent lanes. A more detailed description can be found here.

To access the database:

  • Please print the End-User License Agreement.

  • Sign it, and scan it.

  • Send an email to rfberriel [at] and include the signed EULA as a pdf.

Please note that requests from free email addresses (hotmail, yahoo, gmail, etc.) will be refused. After submission of the EULA, it may take up to a week for you to receive the download link (you will receive a notification by email).

If you use this database in your research, please cite:

    Author  = {Rodrigo F. Berriel and Edilson de Aguiar and Alberto F. de Souza and Thiago Oliveira-Santos},
    Title   = {{Ego-Lane Analysis System (ELAS): Dataset and Algorithms}},
    Journal = {Image and Vision Computing},
    Year    = {2017},
    DOI     = {10.1016/J.IMAVIS.2017.07.005},
    ISSN    = {0262-8856},
You can’t perform that action at this time.