ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question
0

rtabmap_ros: how it works?

asked 2016-10-17 13:54:48 -0500

Marcus Barnet gravatar image

Hi to all,

I'm using the rtabmap_ros package in my project in order to build 3D maps from my recorded bags files. The package works very well and I'm able to obtain accurate maps.

I've read all the documents about the RTAB-Map algorithm and also the articles wrote by Mathieu Labbé. The problem is that I still can't understand how mathematically the closed loop detection works and how the 3D map is built.

Is there any document which I can read in order to understand how the algorithm mathematically works? Can you help me please?

Thank you a lot!

edit retag flag offensive close merge delete

1 Answer

Sort by » oldest newest most voted
3

answered 2016-10-17 16:03:01 -0500

matlabbe gravatar image

updated 2016-10-17 16:03:58 -0500

Hi,

For loop closure detection, see this paper. The code follows pretty much the algorithms shown.

For the 3D map, we simply assemble point clouds of each node in the graph, as described loosely in the paper your referred:

Fig. 9 shows the resulting global map by assembling the RGB-D point clouds from the Kinect using the optimized poses of the graph.

Search Graph-Based SLAM to learn more about the map structure used in RTAB-Map.

cheers

edit flag offensive delete link more

Question Tools

1 follower

Stats

Asked: 2016-10-17 13:54:48 -0500

Seen: 1,047 times

Last updated: Oct 17 '16