Alert icon
We're changing our privacy policy. This stuff matters.  Learn more  Dismiss

Robotics - SLAM and Localization with a Stereo Camera

Loading...

Sign in or sign up now!
15,272
Loading...
Alert icon
Sign in or sign up now!
Alert icon

Uploaded by on May 25, 2007

In this experiment a mobile robot builds a 3D map from stereo images (SLAM with Rao-Blackwellized Particle Filter), and then it uses that map to localize itself within the same room starting at an unknown position. In a related paper we propose an observation likelihood function which does not require to establish correspondences between observations and the map.
Full-resolution videos and papers at:
http://www.isa.uma.es/C10/Temas-Topics/default.aspx
http://www.isa.uma.es/jlblanco

Category:

Entertainment

Tags:

License:

Standard YouTube License

  • likes, 0 dislikes

Link to this comment:

Share to:

Uploader Comments (jlblanco2006)

  • It seems that you use a home made stereo vision system, what's your idea about onboard stereo vision systems? (e.g. w-w-w[dot]focusrobotics[dot]co­m)

  • The camera in this robot is a bumblebee commercial stereo camera, which has given us quite good results by now. We also tried a home-made stereo but had *very* bad results.

    We haven't try the focusrobotics' camera, so I can't tell you the difference with the bumblebee.

  • Is it IMU added or just based on Vision?

  • It is based just on stereo vision. An IMU will improve the results here, but actually it is more necessary in a monocular system (like youtube /watch?v=hgOA7nOy53g or youtube /watch?v=jNuBO0FMRmU).

see all

All Comments (6)

Sign In or Sign Up now to post a comment!
  • What are blue lines?

  • I checked bumblebee. Does it support Linux or you use Windows OS?

  • aaa

  • Great job!

Loading...

0 / 00Unsaved Playlist Return to active list
    1. Your queue is empty. Add videos to your queue using this button:
      or sign in to load a different list.
    Loading...Loading...Saving...
    • Clear all videos from this list
    • Learn more