Tracking of 2D objects via SURF points

Loading...

Sign in or sign up now!
Alert icon
Upgrade to the latest Flash Player for improved playback performance. Upgrade now or more info.
6,447
Loading...
Alert icon
Sign in or sign up now!
Alert icon

Uploaded by on May 15, 2010

This was a sideway attempt to locate 2D objects in an image.
However, it cannot be used in my project of finding game boards, as it suppose to know the content of the game board (unlike the playzone system).
It is based on the computation of the SURF descriptors : for more info, cf http://en.wikipedia.org/wiki/SURF , and on its OpenCV implementation.

Here are the steps : 1. For the model image, compute the grayscale version and compute the SURF features ;
2. Get the input image from the camera, convert it to grayscale ;
3. Compute the SURF features of the grayscale camera frame ;
4. For all features of the model, for all features of the camera frame, determine if they represent the same point (computation of their distance and thresholding) ;
5. Once we have obtained the pairs of associated points, we determine the homography matching all these pairs (using RANSAC or least median squares algorithm) ;
6. Drawing of the projection of the input frame in the illustration frame using this homography.

Here is the caption of the video : * Red circles : SURF features, the radius corresponds to their size (gradient amplitude) * Blue lines : correspondences between the object and the camera frame * Red polygon : when the object is located in the camera frame, drawing of its contour

Category:

Science & Technology

Tags:

License:

Standard YouTube License

  • likes, 0 dislikes

Link to this comment:

Share to:

Uploader Comments (ArnaudRamey)

  • Great job!! I am a curious student.. could you email me or give me a link for the source code. or can you give me detailed algorithm? Thanks!!

  • @neelshah1992 I will release soon this as a ROS stack. Cf ros.org.

  • Well, this is almost out-of-the-box openCV SURF routines, compilated with optimization params of g++. No IPP here. However, the frame rate is not incredibly high, aroud 3-4 FPS an with a CPU load of around 40% (with a AMD 5200+ DualCore and 2 GB of RAM).

see all

All Comments (6)

Sign In or Sign Up now to post a comment!
  • @ArnaudRamey hi, is it possible to have your code, I'm a student now. Thank you very much

  • @ArnaudRamey Thank you sir!! I have never used anything with ROS stack so you will have to explain me that haha.. THANK YOUU

  • amazaing, I thought that finding SURF descriptors and correspondences between them is a very time consuming routine. seems that it isn't so. do you use IPP? what's the average CPU load?

Loading...

Alert icon
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