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

Static 3D Environment Map for localization and path planning

asked 2013-02-07 18:41:39 -0500

Ash420 gravatar image

updated 2014-01-28 17:15:10 -0500

ngrennan gravatar image

Hi, I am working on a robotics project that involves building a 3D static environment map. I am considering Octomap as a possible solution. The 3D map needs to be generated with some process and then utilized for localization and path planning. The 3D map is simple; it is not a complex terrain. The map also contains obstacles. Can someone advise me on how and why to use Octomap? Is it worth the time so that I can incorporate localization and path planning? Please comment on complexity if possible.

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted

answered 2013-02-07 20:59:57 -0500

AHornung gravatar image

You can build a static map with a 3D modelling tool (e.g. Blender, Sketchup) and then convert it to OctoMap, as detailed in these instructions:

There are multiple examples of path planning in OctoMap, check out the following two publications:

Code for 6D localization in octomap is already available in the humanoid_localization stack, with details in the paper Humanoid Robot Localization in Complex Indoor Environments (Video)

You won't get a plug-and-play solution in terms of complexity (you will probably have to read and adjust some code), but in general that has all been done before. As on why to use OctoMap: You will have a memory-efficient 3D occupancy map representation. If you environment is only small and the memory footprint does not matter, you will still have an integrated solution with most code already implemented (map building, raycasting for localization, efficient map access).

edit flag offensive delete link more

Question Tools


Asked: 2013-02-07 18:41:39 -0500

Seen: 2,525 times

Last updated: Feb 07 '13