Hi everyone!

I have been recently through a need for a reliable pose estimation system, in which ar_pose (http://wiki.ros.org/ar_pose) failed to stratify my needs as it depends on the very basic ARToolKit old library.

Moreover, I found aruco_ros (http://wiki.ros.org/aruco_ros) as a good package to begin with, but it was only using a single marker, or double markers. It does not have a visualization system as well.

So, I made my package..

In order to avoid occlusions, I used marker boards (you still have the ability to use a 1x1 marker board), and now it could detect virtually unlimited boards with a very good accuracy.

Nonetheless, it is able to handle many cameras at once, and finally display the result in the rviz (http://wiki.ros.org/rviz).

I would love if you discover things further yourself, so here is the link:

https://github.com/Sahloul/ar_sys

It would only cost you a camera and couple of papers to try, therefore, kindly be asked to try it and let me please know your impression and feedback which is highly appreciated!

_______________________________________________
ros-users mailing list
ros-users@lists.ros.org
http://lists.ros.org/mailman/listinfo/ros-users