ARTSLAM_WRAPPER

artslam_wrapper is a ROS wrapper for ART-SLAM. Right now, it accepts only point clouds as input, and it performs LiDAR SLAM, as intended. It also outputs, periodically, the built 3D map and the estimated positions of the robot (respectively on the /artslam_laser_3d_wrapper/single_cloud and /artslam_laser_3d_wrapper/markers topics).

Requirements

artslam_wrapper requires the following libraries:

• Eigen3
• Boost > 1.65.1
• PCL > 1.10
• OpenCV > 4.0
• g2o
• suitesparse

Moreover, it requires the following packages, along with the corresponding dependencies:

• artslam_laser_3d

Build

artslam_wrapper is built using catkin, although ROS is not mandatory. After having build artslam_laser_3d:

cd catkin_ws/src
git clone https://github.com/MatteoF94/ARTSLAM_WRAPPER.git
cd .. && catkin_make -DCMAKE_BUILD_TYPE=Release

Alternatively to catkin_make, you can use the "catkin build artslam_wrapper" approach, which will automatically build all packages.

Instructions

The file src/artslam_controller.cpp describes in detail how to build your own SLAM system, step by step. If you want to perform SLAM offline (no bags), do the following service call:

rosservice call /artslam_controller/OfflineSLAM