Ask Your Question

3D mapping Cliff Detection [closed]

asked 2021-08-24 03:36:15 -0500

jacklu333333 gravatar image

Hi, I am using rtabmap in my project with Velodyne VLP-16 sensor so that the ground can be identified by the slam algorithm. As mention here that if I want to detect the cliff robustly, then it requires 3D mapping since it can provide groundtracking.

In reality, the stair should be 3 meters in front of the robot. However, in my mapping, as you can see the following.


There is no cliff identified in the map since the lidar didn't san the surface of the stair due to the scanning angle.

To my understanding, the meaning of groudtracking is that the ground will be identified and extend and if the lidar sensor failed to have scan return the cliff will then be identified.

Did I misunderstand anything or misconfig anything, if so please do correct me. Thanks.

Best regards, Jack Lu

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by jacklu333333
close date 2021-09-24 05:31:48.950005

1 Answer

Sort by ยป oldest newest most voted

answered 2021-09-03 15:11:38 -0500

matlabbe gravatar image

Your robot would have to move only on empty cells, and avoid unknown cells. The empty cells seem sparse in your screenshot, you may use point_cloud_assembler node to accumulate multiple velodyne scans while moving and feed the assembled_cloud to rtabmap node instead of the velodyne point cloud. This could help to get denser empty cells in the static map. Parameter Grid/MaxObstacleHeight could be also set to -0.10 so that plane surfaces 10 cm under the ground are not labelled as empty, but obstacles instead.

edit flag offensive delete link more

Question Tools

1 follower


Asked: 2021-08-24 03:36:15 -0500

Seen: 106 times

Last updated: Sep 03 '21