1
$\begingroup$

I need to detect Aruco/Apriltags markers and perform pose estimation, but I've no particular constraint on their physical size. I'm not talking about the number of square (e.g. 4x4, 6x6, etc.), but of the dimension of the markers in millimeters. Which size should I choose, in order to maximise the positive detection rate? Are large markers easier to detect than smaller ones?

I suspect that the full answer will depend non trivially on the distance camera-markers, and possibly on lighting. Assume good lighting and 0.5-1m distance from the markers, if needed.

Any reference is appreciated.

$\endgroup$
2
  • $\begingroup$ A related question has been asked before: robotics.stackexchange.com/q/19901/21531 $\endgroup$ Dec 16 '20 at 5:46
  • $\begingroup$ No, my question is not Aruco vs Apriltag, is about the size of the marker, irrespective of the marker being Aruco or Apriltag. $\endgroup$
    – Rexcirus
    Dec 16 '20 at 10:58
1
$\begingroup$

In general, a larger fiducial will always produce a more robust and accurate measurement. This is because any sub-pixel detection errors will be reduced by the larger baseline. (For example, the detector erroneously measures one corner of the fiducial to be 1 pixel too high. The angular error will be greater if the tag in view is 20 pixels across as opposed to 200).

This is assuming the fiducial isn't so large as to be out of the camera frame. So the exact dimensions of the tag is highly dependent on your physical robot and environment. And if your set up is sufficiently flexible so that the robot must see the tag from very far away and very close, then you must make the trade off where you need the robot to be able to see the tag. Or possibly use multiple tags of different sizes.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.