Re: Does obstacles_detection nodelet assume that the sensor is level?

Posted by ibd on
URL: http://official-rtab-map-forum.206.s1.nabble.com/Does-obstacles-detection-nodelet-assume-that-the-sensor-is-level-tp3485p3492.html

Hi,

Thank you for your help.

The transform device -> base_footprint is correct. I have configured it so that the Tango tablet is sitting on the robot. This is done with a static TF publisher.
Note that this is just a visualization. I am not moving the tango device with respect to the turtlebot. The Tango tablet is sitting on my desk on initialization, and I am moving it by hand. Please ignore the turtlebot model shown on the screenshots.

Here is what the different node(let)s are using for "frame_id" and similar:
rtabmap_ros/pointcloud_to_depthimage:
"fixed_frame_id" = "odom"

rtabmap_ros/obstacles_detection:
"frame_id" = "base_footprint"
"map_frame_id" = "map"

rtabmap:
"frame_id" = "base_footprint"
"odom_frame_id" = "odom"
"map_frame_id" = "map"
"vo_frame_id" = "odom"

I am following the tutorial at this link.

The complete system works fine in the default configuration!
However, just for testing, I wanted to angle the tango device up (turtlebot visualized accidentally). Any other ideas how to fix this? I am making a video to again show the behavior a bit better.