Ask Your Question

Adding meshes to detected transform

asked 2019-08-09 13:19:37 -0600

lauesa gravatar image

OS: Ubuntu 18.04.2 DISTRO: Melodic

I have a small custom mobile robot with a Intel Realsense d435i on it to perform machine vision tasks such as object recognition and grasping. The node I've written to accomplish this classifies objects it comes across on a course and puts out a transform of the object's pose. It publishes the transform as "large_box_1", "large_box_2", etc. When I pull the RobotModel in RVIZ, I'm able to load a robot description I've described in the URDF.

What I'd like to do is take the meshes for the "large_box" I've created, and overlay them on the transforms in RVIZ. This would give better visualization of how accurately the robot is classifying these poses. How would I set this up?

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted

answered 2019-08-12 03:48:57 -0600

If you're detection node can broadcast an unique tf_frame for each object then you can also publish a Marker message containing a mesh resource in that frame. You could also color tint or use transparency so that this model doesn't cover up your original point data.

This method means you can use any STL / DAE file to represent the detected objects, maybe not needed for simple boxes but could be very useful for more complex shapes.

edit flag offensive delete link more



This solution worked very well for my purposes. Thanks for your help!

lauesa gravatar image lauesa  ( 2019-08-12 17:04:41 -0600 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools



Asked: 2019-08-09 13:19:37 -0600

Seen: 36 times

Last updated: Aug 12 '19