Ask Your Question
0

Creating /tf (position) from IMU, 2D/3D SLAM (LiDAR+IMU+Octomap) (no need for high accuracy))

asked 2019-03-20 18:19:19 -0500

cergina gravatar image

Hello, Iam trying to build a simple SLAM system consisted of:

  • RPLiDAR A2 M8
  • Razor IMU 9DOF M0

used packages:

  • octomap_mapping,
  • razor_imu_9dof (razor-pub.launch)
  • rplidar_ros (rplidar.launch)

own packages

  • mctr-132_ros (sends transform of world 0,0,0 and transform of laser which has location 0,0,0 and orientation is quaternion from imu)
  • mctr-132_laser (only converting LaserScan to PointCloud)

System: Ubuntu 16.04, ROS Kinetic

Iam in the phase that I put those points into map and I can visualize it in Rviz, but I only take IMU rotation into account. The position is being published into the /tf topic as 0, 0, 0. I can therefore build map of the surrounding environment. But I'd like to move my device around. So I need a proper /tf with position transform. I was thinking about robot_pose_ekf, but it needs odometry and I have no idea how to create fake odometry from 9dof imu.

Q1: Is it possible to create fake odometry from IMU and if so how?

Many thanks

ps.: I would upload also a photo but 5 points are required...

ps.: I tried google cartographer, hector_localization, rf2o_laser_odometry but I failed to build/run them...

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted
0

answered 2019-10-09 00:53:18 -0500

indraneel gravatar image

updated 2019-10-09 01:00:37 -0500

Hello! I am sorry I am really late to this thread but heres how I would do it, I would also love to hear how you ended up building your SLAM system.

  1. Use a visual odometry software like a laser_scan_matcher (http://wiki.ros.org/laser_scan_matcher) or ICP odometry from rtabmap or hector_mapping (http://wiki.ros.org/hector_mapping) to emulate wheel_odometry data from the RPLiDAR
  2. Fuse the visual odometry with IMU yaw and acceleration using the robot_localization package (This will give you a reliable odometry data source odom->base_link)
  3. Use LiDAR data to build map and publish loop closures using rtabmap/ cartographer to correct the odometry drift (map ->odom)

Thanks!

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower

Stats

Asked: 2019-03-20 18:19:19 -0500

Seen: 183 times

Last updated: Oct 09