multi_sensor_loclization_and_mapping

It is the course work of localization and mapping based on mutiple sensor fusion. Most of the codes are written by the tutors [Qian Ren] and [Yao Ge].

It consists of :

• a mapping pipeline based on graph-based optimization, which utilze RTK, IMU, NDT-based Lidar odometry, and loop closure;
• a localization pipeline based on ESKF, which utlized IMU, NDT-based map matching, motion model and velocity;
• a localization pipeline based on graph optimization, which utlized IMU, NDT-based map matching, motion model and velocity.

Dependencies

1. ROS
2. Eigen
3. PCL
4. g2o
5. ceres solver
6. Sophus
7. Protobuf
8. yaml
9. Geographic
10. Glog

I installed the same version as the third parties in SLAMbook.

Build

catkin_make

Run

Mapping

Run mapping pipeline to build a global map for the environment.

roslaunch lidar_localization lio_mapping.launch

ESKF localization

Run localization pipeline based on ESKF using the built map

roslaunch lidar_localization kitti_localization.launch

Graph-based localization

Run localization pipeline based on graph optimization using the built map

roslaunch lidar_localization lio_localization.launch