search

SteveMacenski / slam_toolbox

Slam Toolbox for lifelong mapping and localization in potentially massive maps with ROS
GNU Lesser General Public License v2.1
1.68k stars 525 forks source link

Two LiDAR setup seems to use wrong frames #462

Closed JuliusSustarevas closed 2 years ago

JuliusSustarevas commented 2 years ago

Required Info:

Steps to reproduce issue

Have 2 lidars, move around , run async mapping

Expected behavior

The maps builds as well as with 1 lidar

Actual behavior

The map builds much more poorly than using 1 lidar. From the weird detections of occupancy in the middle of free space it suggests that scan msg from one scan Is sometimes used as if with another frame id. Running with debug shows that both lidar frames are being read. And occupancy is definitely updated using both. And sometimes correctly. But sumtines it just puts scan readings from one side to another.

Additional information

1Lidar results: image 2 lidars: image

SteveMacenski commented 2 years ago

This project does not support multiple lasers and it is not in the roadmap to do so without contribution from community members

JuliusSustarevas commented 2 years ago

thanks for a quick reply. I'll take a look over xmas if it would be easy to make multi lidars work and make a pull request. I have not checked source yet how difficult this would be.

Could you maybe recommend an approach or recommend what I should look at? Or just what this would entail? Would it be as easy as adding some handler with appropriate tf lookups before the scan message is pushed deeper to scan matching/mapping etc ?

JuliusSustarevas commented 2 years ago

Using ira_laser_tools laserscan_multi_merger solves the issue