mqjinwon / direct_lidar_inertial_odometry

Docker version of Direct LiDAR inertial Odometry

Direct LiDAR-Inertial Odometry: Lightweight LIO with Continuous-Time Motion Correction

This repository is for an easy-to-use library utilizing Docker. We use the following environment.

• docker
  • Ubuntu 20.04
  • ROS noetic

You can find the original instruction file here

We have modified README.md to include instructions for installing Docker images and containers.

1. Build docker image with Dockerfile

Using the image file, you can create a container.

cd docker
docker build -t dlio-noetic:latest .

2. Make container

Using image file, you can make container.

docker run --privileged -it \
           --volume=${DLIO_root}:/root/workspace/src \
           --volume=/tmp/.X11-unix:/tmp/.X11-unix:rw \
           --net=host \
           --ipc=host \
           --name=${docker container name} \
           --env="DISPLAY=$DISPLAY" \
           ${docker image} /bin/bash

You need to replace ${DLIO_root}, ${docker container name}, and ${docker image} with your specific values. An example is provided below.

Example

docker run --privileged -it \
           --volume=/home/jin/Documents/direct_lidar_inertial_odometry:/root/workspace/src \
           --volume=/tmp/.X11-unix:/tmp/.X11-unix:rw \
           --net=host \
           --ipc=host \
           --name=dlio-noetic \
           --env="DISPLAY=$DISPLAY" \
           dlio-noetic:latest /bin/bash

Output

==============Direct Lidar Odometry Noetic Docker Env Ready================
root@jin-940XFG:~/workspace#

3. Build project

❗️ All processes are operated within the container terminal!

source /ros_entrypoint.sh
source /opt/ros/noetic/setup.bash
catkin init
catkin build

4. Execution

❗️ All processes are operated within the container terminal!

roslaunch direct_lidar_inertial_odometry dlio.launch \
  rviz:={true, false} \
  pointcloud_topic:=/robot/lidar \
  imu_topic:=/robot/imu

Services

You can save .pcd and trajectory files.

Save .pcd file

rosservice call /robot/dlio_map/save_pcd LEAF_SIZE SAVE_PATH

Test Data

For your convenience, we provide sample test data here (9 minutes, ~4.2GB).

roslaunch direct_lidar_inertial_odometry dlio.launch

rosbag play dlo_test.bag

Result

Acknowledge

We thank the authors of the direct lidar inertial odometry open-source packages.

Maintainer

• [e-mail]