What is SC-FLOAM?
- A real-time LiDAR SLAM package that integrates FLOAM and ScanContext.
- FLOAM for odometry (i.e., consecutive motion estimation)
- ScanContext for coarse global localization that can deal with big drifts (i.e., place recognition as kidnapped robot problem without initial pose)
- and iSAM2 of GTSAM is used for pose-graph optimization.
- We mainly depend on ROS, Ceres and PCL(for FLOAM), and GTSAM (for pose-graph optimization).
- For the details to install the prerequisites, please follow the FLOAM and LIO-SAM repositiory.
- The below examples are done under ROS melodic (ubuntu 18) and GTSAM version 4.x.
How to use?
- First, install the abovementioned dependencies, and follow below lines.
mkdir -p ~/catkin_scfloam_ws/src cd ~/catkin_scfloam_ws/src git clone https://github.com/cuge1995/SC-FLOAM.git cd ../ catkin_make source ~/catkin_scfloam_ws/devel/setup.bash
- For KITTI (HDL-64 sensor), run using the command (Download KITTI sequence 05 or KITTI sequence 07 first)
##In different terminal roslaunch floam floam_velodyne.launch roslaunch aloam_velodyne floam_velodyne_64.launch rosbag play 2011_09_30_0018.bag
- Thanks to LOAM, FLOAM, SC-ALOAM and LIO-SAM code authors. The major codes in this repository are borrowed from their efforts.