This is a package is a "Adaptive Monte-Carlo Localization in 3D".
It is a particle filter that estimates the localization of a robot moving in a 3D environment without using GPS.
It takes information from an odometry source, point-clouds from an onboard sensor (e.g. laser) and distance measurements from radio-range sensors.
The amcl3d package has been tested under [ROS] Kinetic and Ubuntu 16.04.
If you want more information about the algorithm or use this work in your project, please check and cite the following publication:
Francisco J.Perez-Grau, Fernando Caballero, Antidio Viguria and Anibal Ollero:
To know in more detail the behavior of the package:
Building from Source
To build from source, clone the latest version from this repository into your catkin workspace and compile the package using
cd catkin_workspace/src git clone https://github.com/fada-catec/amcl3d.git cd ../ catkin build
Run the test with
roslaunch ouster_ros os1.launch os1_hostname:=10.5.5.94 replay:=true roslaunch amcl3d amcl3d_test.launch
amcl3d.launch: it contains the start of amcl3d node with a standard configuration of parameters.
roslaunch amcl3d amcl3d.launch
amcl3d_test.launch: this roslaunch allows you to start the RViz with the aforementioned configuration, the amcl3d node, the test-amcl3d node, the bag player and creates a transformation to relate the point-cloud frame of test-amcl3d node with the robot frame of amcl3d node.
roslaunch amcl3d amcl3d_test.launch
Bugs & Feature Requests
Please report bugs and request features using the Issue Tracker.
Supported by ROSIN - ROS-Industrial Focused Technical Projects (FTP).
More information: rosin-project.eu