28
3.7.1.2 2D Lidar map building
ROS SLAM Algorithm：
gmapping：Lidar/scan data and odometer/odom data are required, using PF (Particle Filter).
hector ：Based on the optimized algorithm (solving the least squares problem), the advantages
and disadvantages: no odometer is required, but the radar frame rate requires a very high rate at
40Hz, and the estimated 6-degree-of-freedom pose can adapt to the situation of uneven air or
ground. Initial value has a great influence on the result, so the radar frame rate is required to be
higher.
Cartographer：The cumulative error is lower than the previous two algorithms, and it can
naturally output the covariance matrix and the input term optimized by the back-end.
Lower-cost radars can also have high performance.
Combining these several slam algorithms, the kit uses gmapping algorithm to create slam maps.
The slam_gmapping node receives the sensor_msgs/LaserScan message and builds a map
(nav_msgs/OccupancyGrid). The map can be viewed through ROS topic or service.
Gmapping topics and service
Gmapping Coordinate transformation
Name Type
TF transformation Description
Description
coordinate system,
datum and distance
measurement related
frame conversion
tfMessage
base_link-><laser_link>
odom->base_link
map->odom
Topics subscribe
Necessary TF transformation
Necessary TF transformation
Topic publish
Service
tf
scan sensor_msgs/LaserScan
nav_msgs / MapMetaData
Publish Meta data
Publish map raster data
Call this service to get map data
Publish the estimation of
robot pose distribution entropy
nav_msgs / OccupancyGrid
std_msgs / Float64
nav_msgs / GetMap
Lidar scanning data
map_metadata
map
entropy
dynamic_map
Direct conversion between
chassis coordinate system
and lidar coordinate system
The transformation between
the mileage coordinate
system and the chassis
coordinate system, generally
issued by the odometer
The transformation between
the map coordinate system
and the odometer coordinate
system to estimate the pose
of the robot in the map
To Next Page
Loading...
