lmono

LMONO-Fusion: An Online SLAM System based on LiDAR-Monocular Camera Sensor Fusion

This is a framework of LiDAR-monocular camera fusion system. Visual information from monocular camera assist in scene recognization in LiDAR odometry and dense mapping. Automatic calibration between LiDAR and a single camera is provided as well. The pre-print version of our paper is available [here]().

• laser odometry
• scene recognition/ loop detection
• automatic and real-time calibration (Code of this part is not fully uploaded)
• 3D dense map with color information

More code is coming...

• ground segmentation
• communication interface between LiDAR and Camera
• speeding up

Feb, 2022 This work is already submitted to TVT.

1. Prerequisites

1.1 Ubuntu and ROS

• Ubuntu 64-bit 18.04

• ROS Melodic. ROS INSTALLATION

1.2 Ceres Solver

• Ceres. Ceres INSTALLATION

1.3 PCL

• PCL PCL INSTALLATION

2. Build

2.1 Clone repository

 cd ~/catkin_ws/src
 git clone https://github.com/bobocode/lmono.git
 cd ..
 catkin_make
 source ~/catkin_ws/devel/setup.bash

2.2 Download dataset and test rosbag

• Download KITTI sequence

• To generate rosbag file of kitti dataset, you may use the tools

  roslaunch aloam_velodyne kitti_helper.launch
  

2.3 Launch ROS in different terminals

• Launch ALOAM to gain LiDAR measurements

  roslaunch aloam_velodyne aloam_velodyne_HDL_64.launch
  

or

   roslaunch aloam_velodyne aloam_velodyne_HDL_32.launch

• Launch Estimator Node

   roslaunch monolio kitti_estimator_xx.launch
  

• Rosbag Play

  rosbag play xxx.bag
  

• If you want to open loop detection

   roslaunch monolio kitti_loop_xx.launch
  

• If you want to open fusion mapping

   roslaunch monolio kitti_map_00.launch
  

3.Acknowledgements

Thanks to A-LOAM and VINS-MONO code authors. The major codes in this repository are borrowed from their efforts.