fast_lio-sam_loop-gps

Fast-lio with loop factor and GPS factor for back-end optimization. lio and back-end implementation are moved to "include/core" for better readability.

0. Features

GPS information is integrated into fast_lio-sam_loop-gps to build a consistent map in a large-scale environment. We mainly follow the implementation in LIO-SAM-6axis about the system init when we use a 6-axis imu. In addition, we add a two-manual parameter (manual_gps_init+manual_init_yaw) when you are sure about the transformation between the imu and ENU coordinate system. The rebuild ikd-tree modes can still be chosen when the loop closes（correct_fe_en parameter in config file）.

The system computing efficiency is better than LIO-SAM-6axis for incrementally maintaining about ikd-tree map in the odometry. So, using correct_fe_en == false is better for most cases.

1. Prerequisites

1.0 gcc and g++

gcc and g++ 7.5.0 are tested OK.

1.1 Ubuntu and ROS

Ubuntu >= 18.04

ROS >= Melodic. ROS Installation

1.2. PCL && Eigen

PCL >= 1.8, Follow PCL Installation.

Eigen >= 3.3.3, Follow Eigen Installation.

1.3. livox_ros_driver

Follow livox_ros_driver Installation.

1.4. gtsam

Use 4.0.0 version in ubuntu 18.04.

1.5. rviz_satellite (optional)

Follow the "rviz_satellite" bag in LIO-SAM-6axis for better visualization.

2. Build

Clone the repository and catkin_make:

    cd ~/$A_ROS_DIR$/src
    git clone https://github.com/Hero941215/fast_lio-sam_loop-gps
    cd fast_lio-sam_loop-gps
    git submodule update --init
    cd ../..
    catkin_make
    source devel/setup.bash

• Remember to source the livox_ros_driver before build (follow 1.3 livox_ros_driver)

3. Run

3.1. Download Dataset (vlp-16):

rosbag play XXX.bag

3.2. Run SLAM system:

roslaunch fast_lio_sam_loop mapping_velodyne_gps.launch

3.3. Save map:

rosservice call /service/save_map

4. Acknowledgments

Thanks for LOAM(J. Zhang and S. Singh. LOAM: Lidar Odometry and Mapping in Real-time), FAST-LIO2，FAST_LIO_SAM, FAST_LIO_LC, LIO-SAM, LIO-SAM-6axis.