cartographer_ros
cartographer_ros copied to clipboard
How much pose accuracy can be got?
Hi, I'm working with the cartographer 2D SLAM and pure localization. The mobile robot is equipped with two velodyne sensors and odometry. The environment is a closed room, the rectangular map which also allows the robot recognize its heading. My question is by using probability map or tsdf, how much pose accuracy can be obtained? The velodyne scan data seems to give +- 10 mm standard deviation and the resolution of map is 50 mm. Is it possible to get position accuracy within 5 mm even though lidar sensors has +-10 mm?