## repository: https://github.com/RobustFieldAutonomyLab/turtlebot_exploration_3d.git
<<StackHeader(turtlebot_exploration_3d)>>

{{attachment:turtlebot_exploration_icon.png||width="70%"}}


<<TableOfContents>>
== What is it ? ==
This is an ROS implementation of infomation-theoretic exploration using turtlebot with a RGBD camera (e.g. Kinect). It is designed for autonomous mapping of indoor office-like environments (flat terrain). All the computation is performed on the turtlebot laptop and intermediate results can be viewed from remote PC. The output consist of both 2D and 3D Octomap (.ot) file and saved on the turtlebot laptop.

== Papers ==
If you find the package useful, please consider citing the following papers:

{{{
@inproceedings{bai2016information,
  title={Information-theoretic exploration with Bayesian optimization},
  author={Bai, Shi and Wang, Jinkun and Chen, Fanfei and Englot, Brendan},
  booktitle={Intelligent Robots and Systems (IROS), 2016 IEEE/RSJ International Conference on},
  pages={1816--1822},
  year={2016},
  organization={IEEE}
}
}}}

{{{
@inproceedings{bai2015inference,
  title={Inference-Enabled Information-Theoretic Exploration of Continuous Action Spaces},
  author={Bai, S. and Wang, J. and Doherty, K. and Englot, B.},
  booktitle={International Symposium of Robotics Research},
  year={2015}
}
}}}


== Download and Install ==

=== Dependencies ===
Install octomap_ros and rviz plugins.
{{{
sudo apt-get install ros-indigo-octomap*
}}}


=== from source ===

Please download the source code from: 
{{{
git clone https://github.com/RobustFieldAutonomyLab/turtlebot_exploration_3d.git
}}}

then
{{{
catkin_make
}}}

=== binary form ===

{{{
sudo apt-get update
}}}

{{{
sudo apt-get install ros-indigo-turtlebot-exploration-3d
}}}


== Setup ==

Please follow the turtlebot network [[http://wiki.ros.org/turtlebot/Tutorials/indigo/Network%20Configuration|configuration]] to setup network between turtlebot and remote PC.


== Run ==

=== From turtlebot laptop ===
{{{
$ roslaunch turtlebot_exploration_3d minimal_explo.launch
$ roslaunch turtlebot_exploration_3d turtlebot_gmapping.launch
$ rosrun turtlebot_exploration_3d turtlebot_exploration_3d
}}}

=== From Remote PC ===
{{{
roslaunch turtlebot_exploration_3d exploration_rviz.launch
}}}

=== Topics ===

{{{
#!clearsilver CS/NodeAPI
no_header=True
sub {
   0{
    name = tf
    type = tf/StampedTransform
    desc = looking for transformation between /map and /camera_rgb_frame
    }
   1{
    name = /camera/depth_registered/points
    type = sensor_msgs/PointCloud2
    desc = point cloud from Kinect sensor, can remap to a different topic, however have to be similar to Kinect. Otherwise need to update the sensor model in the source code.
    }
  }
pub {
   0{
    name = octomap_3d
    type = octomap_msgs/Octomap
    desc = The octomap generated by this node, published only after each observation.
   }
  }
}}}


== Video ==

A brief demo showing how it works:(video played 5X faster):

<<Youtube(ocZOaySmwHU)>>

== Organization ==

RFAL - 
Stevens Institute of Technology


## AUTOGENERATED DON'T DELETE
## CategoryPackage