[ros-users] ar_pose orientation accuracy

Nikhil Karnad karnad at cs.umn.edu
Mon Nov 8 16:30:46 UTC 2010


Steven,

In my experience, there are two things we can do to make ARToolkit tracking
improve, but I have not had the time to test them.

First is to use arGetTransMatCont(...) which reduces jitter by using
previous frame estimates in the optimization process (see
examples/simpleTest2.c). The second is to edit 'config.h' and increase the
pattern normalization size parameters #defined as AR_PATT_SIZE_X and
AR_PATT_SIZE_Y (and recompiling). This might come at the cost of slowing
things down, but that's the trade-off. I suggest using a filter on top of
the per-frame readings to smooth it out.

@misc{kato2009inside,
title={{Inside ARToolKit}},
author={Kato, H.},
year={2009},
month={03}-19}. http://www.hifl.washington.edu/artoolkit/documentation}
}

As far as the magnitude of the jitter, I found a reference that might be
useful:

@conference{malbezin2002measuring,
title={{Measuring ARTootKit accuracy in long distance tracking
experiments}},
author={Malbezin, P. and Piekarski, W. and Thomas, B.H.},
booktitle={Augmented Reality Toolkit, The First IEEE International
Workshop},
pages={2},
isbn={0780376803},
year={2002},
organization={IEEE}
}

Best,
Nikhil

----
Nikhil Karnad
Department of Computer Science and Engineering
University of Minnesota, Twin-Cities
http://www-users.cs.umn.edu/~karnad/

----
On Mon, Nov 8, 2010 at 7:53 AM, Steven Bellens <
steven.bellens at mech.kuleuven.be> wrote:

> 2010/11/8 Steven Bellens <steven.bellens at mech.kuleuven.be>:
> > Hi,
> >
> > I'm experimenting a bit with the ar_pose package. I'm using a single
> > fixed camera to track a moving marker. To verify the estimation
> > accuracy, I just leave the marker fixed and I've plotted position and
> > orientation estimates. The position estimates are pretty much
> > constant, but the orientation estimates are oscilating a lot (see
> > appendix), and apparently always between two values. Is this because
> > of the bad capability to estimate that orientation or can this be
> > caused by the environment conditions (light - set-up - distance to
> > marker)?
>
> For clarity, plotted are the 4 components of a unit quaternion.
>
> >
> > regards,
> >
> > Steven
> >
> _______________________________________________
> ros-users mailing list
> ros-users at code.ros.org
> https://code.ros.org/mailman/listinfo/ros-users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ros.org/pipermail/ros-users/attachments/20101108/f3221346/attachment-0003.html>


More information about the ros-users mailing list