💾 l3xz_mapping

This repository includes the ROS-based explorative mapping stack of L3X-Z for ELROB 2022.

The stack is tested under ROS1 Noetic Ninjemys. It can do mapping using RTAB-Map and a simple known-position approach. Also data logging is possible.

Installation

Clone the repository and go to host root:

git clone --recursive https://github.com/107-systems/l3xz_mapping
cd l3xz_mapping/host_ws

Build using catkin:

catkin_make install

There is also a frontend available for the mapping stack.

Start mapping

RTAB-Map

roslaunch l3xz_mapping l3xz_mapping.launch

Known-position approach

roslaunch l3xz_mapping knownposition_lidar.launch

Data logging

roslaunch odom_recorder.launch
roslaunch thermal_recorder.launch
roslaunch plotter_grid.launch

Nodes

transform_odom

Creates an odometry message from tf data.

Published Topics

Default Name	Type
/odom_slam	nav_msgs/Odometry

Parameters

Name	Default	Description
odometry_out	/odom_slam	Odometry topic name
parent_frame	/map	Parent frame of interesting tf tree part
child_frame	/base_link	Child frame of interesting tf tree part

plotter

Plots robot track and OPIs according to ELROB rules into a map. The input map can be an image or data from a topic of type nav_msgs/OccupancyGrid. The OPIs to be plotted and the robot track are subscribed from a topic of type l3xz_mapping/Track.

Parameters

Name	Default	Description
from_grid	false	Get map topic from OccupancyGrid
img_src	/home/log/map.png	Map image name
img_dest	/home/log/map_dest.png	Destination map image name
meters_per_pixel	0.1	Resolution of map image
bearing_deg	0.0	Robot initial bearing
zero_px	[100, 100]	Robot initial position
robot_track	/l3xz/odom_recorder/track	Track topic
opi_tracks	[/l3xz/thermal_recorder/track, /l3xz/radiation_recorder/track]	OPI topics
grid_topic	/grid	Grid map

recorder

Records waypoints according to ELROB log format with Unix timestamp and WSG84 coordinates with images of the artifacts. Also a track containing all waypoints can be published (e.g. for plotting). All the logs are based on an initial startpoint.

Service clients

Default Name	Purpose
recorder/set_startpoint	Set initial robot position. Needs to be done only once and is necessary to start recording.
recorder/set_waypoint	Record a new waypoint.

Published Topipcs

Default Name	Type
recorder/track	[l3xz_mapping/Track]

Parameters

Name	Default	Description
logpath	/home/l3xz/log.txt	Path for logfile
artifacts	[]	List of artifact images
track_publishing_rate	1	Period for publishing track rate
publish_track	true	Enable publishing of track

knownposition_node

For some tasks a complete map of an area is not desired. Therefore, it is a good idea to use a robust mapping from known position approach instead of SLAM to observe the area around the robot to master autonomous navigation tasks. This node demonstrates this approach using the data of one 2D-Lidar. All features of the implemented algorithms, e.g. overlaying multiple datasets, image postprocessing or transforming using tf are not demonstrated yet.

Subscribed Topipcs

Default Name	Type
/lidar	sensor_msgs/Odometry
/odom	nav_msgs/Odometry

Published Topipcs

Default Name	Type
/grid	nav_msgs/OccupancyGrid

Parameters

Name	Default	Description
pixel_x	600	Columns of the gridmap
pixel_y	600	Rows of the gridmap
resolution	0.1	Resolution of the gridmap in meters
rate_hz	5	Period for publishing the gridmap
grid_topic	/grid	Name of the gridmap
odometry_topic	/odometry	Name of the odometry topic
show	true	Show map in OpenCV window (like above)