ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

RTAB-Map is a Graph-SLAM approach, while GMapping is a particle filter. With 2D lidar, they will produce similar maps. In term of accuracy, I will let you test both in your environment as I cannot be not the most objective ;) but in indoor small environments, the resulting maps would be accurately similar (maybe gmapping is easier to setup at first as its default parameters target directly 2d lidar).

RTAB-Map is a Graph-SLAM approach, while GMapping is a particle filter. With 2D lidar, they will produce similar maps. In term of accuracy, I will let you test both in your environment as I cannot be not the most objective ;) but in indoor small environments, the resulting maps would be accurately similar (maybe gmapping is easier to setup at first as its default parameters target directly 2d lidar).

RTAB-Map is a Graph-SLAM approach, while GMapping is a particle filter. With 2D lidar, they will produce similar maps. In term of accuracy, I will let you test both in your environment as I cannot be the most objective ;) ;). I did some comparisons in this paper (see figure 13), but in as a disclaimer, I didn't tune gmapping as I did for rtabmap for that dataset. In indoor small environments, the resulting maps trajectories would be accurately similar (maybe gmapping is easier to setup at first as its default parameters target directly 2d lidar).

RTAB-Map is a Graph-SLAM approach, while GMapping is a particle filter. With 2D lidar, they will produce similar maps. In term of accuracy, I will let you test both in your environment as I cannot be the most objective ;). I did some comparisons in this paper (see figure 13), but as a disclaimer, I didn't tune gmapping as I did for rtabmap for that dataset. dataset (probably with some gmapping tuning, results would have been better than with the default parameters, same for the other approaches). In indoor small environments, the resulting trajectories would be accurately similar (maybe gmapping is easier to setup at first as its default parameters target directly 2d lidar).