Quick Start Guide

Updated 3 May, 2016 -- Steven Waslander

1. Environment setup

MCPTAM is maintained for Ubuntu 14.04 and ROS Indigo only, although its REL 0.1.3 is known to work in Ubuntu 12.04 and ROS Hydro as well. The following instructions assume you are working in Ubuntu 14.04 and ROS Indigo.

The necessary ROS dependencies include: roscpp, image_transport, sensor_msgs, cv_bridge, geometry_msgs, tf, message_filters, std_msgs, std_srvs, visualization_msgs, dynamic_reconfigure, pcl_ros, message_generation, libg2o.

The non-standard ROS packages can be installed with apt-get:

sudo apt-get install ros-indigo-image-transport ros-indigo-image-transport-plugins ros-indigo-pcl-ros ros-indigo-libg2o

Although not part of the default install, if you would like to use USB cameras that run using the uvc_camera node:

sudo apt-get install ros-indigo-uvc-camera

The list of non-ROS dependencies include: boost, suitesparse, blas, lapack, toon, openCV, libcvd, openGL, freeglut3 and gvars3. These dependencies are all managed as part of the script installation below.

2. MCPTAM Script installation

As part of our Travis CI integration, it is now possible to install MCPTAM and its dependencies with one script. Note that the script assumes no prior install of MCPTAM, and that a properly configured catkin workspace directory is provided. If you don't have a catkin workspace, follow these instructions to set it up.

To use the script for installation:

1. Move to the location you've created your catkin workspace (usually /home/username/catkin_ws)

    cd ~/catkin_ws
   

2. Grab the mcptam install script from the repository

    wget https://github.com/wavelab/mcptam/raw/master/scripts/mcptam_install.bash
   

3. Convert the script to be executable

    chmod a+x mcptam_install.bash
   

4 . And finally run it specifying the catkin_ws directory as an argument. If run from within catkin_ws:

    ./mcptam_install.bash .

or if run from elsewhere:

    ./mcptam_install.bash /home/username/catkin_ws

3. Running MCPTAM on Recorded (bagged) Data

It is possible to run MCPTAM on bagged data to get a sense for its capabilities. A three camera bagged dataset is available at http://wavelab.uwaterloo.ca/sharedata/TravisCI/Flight_Arena.bag. The 431 MB data file covers a one minute test in our motion capture studio, with the following data recorded:

• Camera: 3x ximea cameras, captured in greyscale, at 900x600 pixel resolution (60Hz each).
• IMU: Microstain 3DM-GX3-25 (125 Hz).

To test your installation on bagged data,

1. Start roscore.

    roscore
   

2. Open the terminal to make ros use the clock from the bag file, and not real time from your machine.

    rosparam set use_sim_time true
   

3. Play the bag file from wherever it is saved,

    rosbag play --clock <bag_file.bag>   
   

4. Launch MCPTAM from a terminal with the right group file. Example group files are in the groups folder of MCPTAM source, need to match the cameras.

    roslaunch mcptam mcptam.launch group_name:=cameras_1,2,3
   

   Note: If mcptam fails to launch its GUI at this point, check the mcptam.launch file parameters and confirm that the image_transport parameter 'value' is set to 'compressed'.

5. Launch rviz to see the map points.

    rosrun rviz rviz -d <path to mcptam.rviz>
   

   Note: The mcptam.rviz file is located in the mcptam/launch directory.

6. Go back to MCPTAM window and hit 'Spacebar' to view map points.