A ROS node to perform a probabilistic 3-D/6-DOF localization system for mobile robots with 3-D LIDAR(s). It implements pointcloud based Monte Carlo localization that uses a reference pointcloud as a map.
BSD 3-Clause "New" or "Revised" License
480
stars
119
forks
source link
Robot movement needs to follow ROS frame convention? #286
Hi @at-wat ,
In my case, the odometry estimation comes from my VIO, hence, in the local odom frame, the forward movement happens along the y-axis as in this picture:
Do we need to transform the local odom frame to match the global map frame followed ROS convention?
When I directly use the VIO as odometry sources, the result looks like this:
The pointcloud comes from a TOF camera and it can not align with the map.
I would like to ask if there are some frames alignments (between map, odometry, base_link, pointcloud, IMU) that need to be done before starting the filter?
Thanks for your help!
Sorry for delayed reply.
Current implementation uses IMU acceleration vector to obtain the ground direction.
So, I guess you need to adjust IMU frame relative to base_link in your case.
Hi @at-wat , In my case, the odometry estimation comes from my VIO, hence, in the local odom frame, the forward movement happens along the y-axis as in this picture:
Do we need to transform the local odom frame to match the global map frame followed ROS convention?
When I directly use the VIO as odometry sources, the result looks like this:
The pointcloud comes from a TOF camera and it can not align with the map.
I would like to ask if there are some frames alignments (between map, odometry, base_link, pointcloud, IMU) that need to be done before starting the filter?
Thanks for your help!