Dear sir,
We're using the "2.3 Kinect + 2D laser" example with RTAB-Map's own icp_odometry node. We're working with a 3DoF robot and only need a 2D occupancy grid map.
Can you please explain (or attach a link) as to how the loop closing is done? Is it done by the camera or ICP? If done by the camera, is the occupancy grid map fixed after a loop closure?
We're using:
<param name="Reg/Force3DoF" value="true" />
<param name="Optimizer/Slam2D" value="true" />
Thank you, Yoni