LIME is an open source 3D LiDAR-based localisation framework. This framework enables localisation in a premapped environment. Prior mapping can be achieved by using packages like loam_velodyne and LeGO_LOAM. Wheel odometry is used to provide high frequency, low accuracy updates which are corrected by low frequency, high accuracy ICP corrections.

video

This package is built on top of the ROS ecosystem.

This package has been tested on Ubuntu 16.04 & ROS kinetic.

premapped_localization depends on the following libraries:

• PCL : Installation instructions
• libpointmatcher : Installation instructions

Installing this package this package:

cd ~/catkin_ws/src
git clone https://github.com/ShreyanshDarshan/premapped_localization.git

cd ~/catkin_ws/
catkin_make

In a terminal execute:

cd ~/catkin_ws/
source devel/setup.bash
roslaunch icp setup.launch

If using a bagfile, in a different terminal, run:

rosbag play <path to bagfile>/<bagfile_name>.bag

Make sure that the encoder data is being published at the topic /encoders of message type geometry_msgs/Twist and is of the following format:

• twist.linear.x -> left wheel encoder data
• twist.linear.y -> right wheel encoder data