Robot fails to localize with error "The origin for the sensor at (0.06, 0.02, 0.30) is out of map bounds. So, the costmap cannot raytrace for it."

Vikranj

December 30, 2019 01:59

I am using D435 and T265 on Nvidia Jetson Tx2 to navigate Trutlebot2

I am creating a 2d map of the environment using the occupancy package (https://github.com/IntelRealSense/realsense-ros/tree/occupancy-mapping/occupancy). I am using the following command to save the map.
```
rosrun map_server map_saver map:=/occupancy –f my_map_1
```
I then use depthimage_to_laserscan package to convert D435 depth frames to laser scan.
I copied the R200 costmap setting which is in https://github.com/turtlebot/turtlebot_apps/tree/indigo/turtlebot_navigation/param.
I modified amcl_demo launch file in https://github.com/turtlebot/turtlebot_apps/tree/indigo/turtlebot_navigation/launch to work with D435.

I keep getting this error and the robot fails to localize. I am not sure how to fix the issue.

Also, will it be better if I use D415 instead of D435?

Is there any code samples for the demo in this video.

https://www.youtube.com/watch?v=62vm0_RZ1nU&feature=youtu.be&t=1036

Thanks.

0

Comments

1 comment

MartyG

December 30, 2019 07:05
The official FAQ for the T265 suggests that the D415 model may be the optimum pairing for best depth quality (presumably as the D415 has around half the RMS error (error over distance) of the D435 models). A D435 model may be best for you though if your robot is going to be fast-moving, as those models have a faster 'global shutter', whilst the D415 has a slower 'rolling shutter'. The D415 can cope with human walking pace though.

In 2017 a complete hardware and software guide for constructing a RealSense powered Turtlebot was published. It was originally designed for the older R200 and ZR300 cameras, though the author said that the guide was adaptable for the 400 Series cameras, which were launched 2 months after the guide's last update.

https://software.intel.com/en-us/articles/build-an-autonomous-mobile-robot-with-the-intel-realsense-camera-ros-and-sawr

In the guide, the author recommends using a common ROS module called move_base for maintaining cost maps. Use of cost maps with this module are discussed on pages 40 to 50 of the paper in the PDF lnk below.

https://www.utupub.fi/bitstream/handle/10024/147661/VSLAM_and_Navigation_System_of_Unmanned_Ground_Vehicle_Based_on_RGB-D_Camera.pdf
0

Comment actions Permalink

Please sign in to leave a comment.