Skip to content

Building map

Overview

Purpose: Use the mapping program to obtain a map (GlobalMap.pcd)

LIO-SAM-6AXIS

lio-sam-6axis

step-1 Calibrate the sensor (the initial parameters have been set, you can skip it. If the mapping effect is not good, please recalibrate it)

step-2 Record ros2 bag

NOTE Check the topic output before recording. During the recording process, the vehicle should be driven as smoothly and slowly as possible. It is recommended to loop your map scene.

  • Start autoware first:./autoware.sh Note
Method 1 ：cd  ~/pixkit/pix/ros2_bag
                  ros2 bag record $(cat rosbag.txt)
Method 2 ：cd  ~/pixkit/pix/ros2_bag（by yourself）
                  ros2 bag record  /lidar_topic /gnss_topic /imu_topic 
                  Here   /lidar_topic：lidar topic
                        /gnss_topic ：gbss topic
                        /imu_topic  : imu topic

step-3 Modify the config file

run.launch under launch file:

import os
from ament_index_python.packages import get_package_share_directory
XXX
XXX

def generate_launch_description():

    share_dir = get_package_share_directory('lio_sam')
    parameter_file = LaunchConfiguration('params_file')
    xacro_path = os.path.join(share_dir, 'config', 'robot.urdf.xacro')
    rviz_config_file = os.path.join(share_dir, 'config', 'rviz2.rviz')

    params_declare = DeclareLaunchArgument(
        'params_file',
        default_value=os.path.join(
            share_dir, 'config', 'params_rs16.yaml'),
        description='FPath to the ROS2 parameters file to use.')

params_rs16.yaml：Read the configuration file under config. If you need to change it, please pay attention here

  • You need to modify the topic configured for your sensor lio-sam-topic
parameter msg type describe
pointCloudTopic sensor_msgs/Pointcloud2 The topic where you recorded the LiDAR point cloud
imuTopic sensor_msgs/Imu The topic where you recorded the IMU
odomTopic nav_msgs/Odometry IMU odometer topic, no modification required
gpsTopic nav_msgs/NavSatFix The positioning topic you recorded for GNSS
  • If GNSS is used for orientation initialization, it needs to beuseImuHeadingInitializationset to true， gps-params

step-4: Run the mapping program

Note Must end autoware

Run the map launch file under mapping_ws

source install/setup.bash
ros2 launch lio_sam run.launch.py

step-5: play ros2 bag

ros2 bag play  [your_ros2_bag] --clock 
your_ros2_bag： Play ros2 bag, the following operations are required after playing completed ros2 bag

step-6: Save the map

Under the mapping_ws file：

source install/setup.bash
ros2 service call /lio_sam/save_map  lio_sam/srv/SaveMap

The saved map is in LOAM under Downloads and will be used

name describe
GlobalMap.pcd Point cloud map，import Vector Map Builder to draw, if using GNSS, the direction has been converted to ENU direction

The origin WGS84 coordinates of the point cloud map, used for GNSS positioning

It is recommended to save the information in the terminal for easy use in GNSS positioning. map-origin

Code git

prerequest

  • Ubuntu 22.04
  • ROS2 Humble

step-1: Build a drawing workspace

Create a workspace

mkdir ~/mapping_ws
cd ~/mapping_ws
mkdir src
cd src

step-2 clone

git clone https://github.com/pixmoving-moveit/LIO-SAM-ROS2-PIXKIT # clone

step-3 Install dependencies

sudo apt install ros-humble-perception-pcl \
       ros-humble-pcl-msgs \
       ros-humble-vision-opencv \
       ros-humble-xacro
# Add GTSAM-PPA
sudo add-apt-repository ppa:borglab/gtsam-release-4.1
sudo apt install libgtsam-dev libgtsam-unstable-dev

step-4 colcon build package

cd ~/mapping_ws
colcon build --symlink-install --cmake-args -DCMAKE_BUILD_TYPE=Release

LiDAR Parameters

lidar-params

parameter describe
sensor LiDAR type, optional velodyne, ouster, livox, robosense
N_SCAN Point cloud channel number, 16-line laser radar is 16
Horizon_SCAN LiDAR lateral resolution, (Velodyne: 1800, Ouster: 512, 1024, 2048, Livox Horizon: 4000)
downsampleRate Downsample rate, let N_SCAN/downsampleRate=16
lidarMinRange Minimum point cloud distance, default is 1.5
lidarMaxRange Maximum point cloud distance, default is 1000.0
  • IMU internal parameters, refer to IMU Calibrationto obtain IMU internal parameters

IMU intrinsics

parameter describe
imuAccNoise 3-axis accelerometer white noise mean (m/s^s)
imuGyrNoise 3-axis gyro white noise mean (rad/s)
imuAccBiasN 3-axis accelerometer zero bias mean (m/s^s)
imuGyrBiasN 3-axis gyro bias mean (rad/s)
imuGravity Local gravity acceleration (m/s^2)
imuRPYWeight Angle weight, default 0.01

IMU extrinsic parameters can be measured by [LiDAR and IMU extrinsic calibration] or manually measured with a tape measure (the coordinate system is the right-hand rule) When measuring manually： - extrinsicTrans:manual measurement with a tape measure - extrinsicRot: the unit matrix is ​​sufficient

IMU extrinsics

parameter describe
extrinsicTrans Translation Matrix
extrinsicRot Rotation Matrix

References