Hi.
I made the construction of the map using the lidar and using the camera. I noticed that the lidar map is more accurate than the camera map. Can I somehow improve the map from the camera.
Thanks for your reply.
Make sure TF of the camera is correctly oriented. For example, if the camera in reality is tilting down, but not in URDF, the walls won't appear straight in base frame, causing thick walls. This can be debugged in RVIZ, set global frame to your base frame, show the point cloud of the camera, and make sure the ground in the cloud is perfectly aligned with the RVIZ grid.