How to generate an 2D occupancy grid from a cloud points from an RGBD camera

I am using Ubuntu 20.04 + ros noetic + opencv 4.4 + orb slam 3 + orb slam wrapper by thien64 This allows me to generate cloud points from the image sequence (tested on TUM rgbd dataset). I want to turn these cloud points into a 2d occupancy grid. A plus would be the use of octomap but i can't seem to find any indication online or clear tutorial.

I don't have much background with ros or mapping any help counts.

Asked by Iscelia on 2023-04-25 10:15:28 UTC

Comments

Answers

The pointcloud_to_laserscan package converts a 3D Point Cloud into a 2D laser scan, which can be easily used for 2d occupancy grid. This is useful for making devices like the Kinect appear like a laser scanner for 2D-based algorithms (e.g. laser-based SLAM).

ps: I am a little confused for you turning cloud points into a 2d occupancy grid with octomap, as octomap is mainly used for creating 3d maps.

Asked by MichaelShuo on 2023-06-27 22:04:37 UTC

Comments

Thanks, the idea behind the use of octomap is that id we already have a 3D map we can turn it easily into a 2D map

Asked by Iscelia on 2023-06-30 05:04:12 UTC