slam_gmapping occupied cells

asked 2014-01-14 04:37:41 -0500

acp gravatar image

Dear all.

I have my LaserScan data, and I have some issues,

*when I run slam_gmapping, it does not mark as occupied cells some obstacles detected by the laser data.

*When gmapping has detected an empty space an suddenly the robot laser detects an occupied space in the same place where it was empty, gmapping doesnt detect the new occupied cells.

I have been playing with:

map_update_interval

temporalUpdate

but no results, do you have some advice?

In advance, thank you :)

edit retag flag offensive close merge delete

Comments

1

Are you talking about an obstacle suddenly appearing in reality? That would violate the static map assumption and thus gmapping might be quite slow to update that.

dornhege gravatar imagedornhege ( 2014-01-14 04:50:03 -0500 )edit

It is in the process of making the map using gmapping, I am using two sensors to make the map. While the robot is travelling, one sensor field of view detect some obstacles, and the other one not, and when the robot, for instance turn around, the other sensor detect other obstacles, but this new obstacles are not detected by gmapping.

acp gravatar imageacp ( 2014-01-14 20:39:33 -0500 )edit

But I am also taking about the first reading of a laser in the process of map making, some obstacles are detected and other are not by gmapping.

acp gravatar imageacp ( 2014-01-14 20:53:57 -0500 )edit