Vision for Robots 1 of n: Setup

This is going to be a mini-series of posts that helps you build a practical computer vision application for your robot. We will be using the ROS robotics framework and OpenCV computer vision library.

To complete this you will need:

• A computer
• Some programming experience
• Optional USB Webcam

The first step is going to be to install ROS. On Ubuntu, if you don't like reading instructions, you can try copy and pasting this and it might work. It may also cause your computer to catch on fire, but these are the risks you take when you refuse to read the instructions.

sudo apt-get install build-essential python-yaml cmake subversion wget
wget --no-check-certificate http://ros.org/rosinstall -O ~/rosinstall
chmod 755 ~/rosinstall
~/rosinstall ~/ros http://ros.org/rosinstalls/boxturtle_base.rosinstall
echo "source ~/ros/setup.sh" >> ~/.bashrc
. ~/.bashrc
rosdep install ros
rosmake install ros
rosdep install image_view
rosmake install image_view

Once you manage to get ROS installed you may want to spend some time going through the tutorials. They are well worth your time.

If you have a USB webcam that supports the uvc video class you can install one of the drivers.

cd ~/ros/stacks
svn co https://bosch-ros-pkg.svn.sourceforge.net/svnroot/bosch-ros-pkg
cd ~/ros/stacks/bosch-ros-pkg/trunk/bosch_drivers/usb_cam
rosdep install usb_cam
rosmake usb_cam

To test the usb_cam driver create a launch file directory.

mkdir ~/ros/launch

Next, create a launch file called ~/ros/launch/usb_cam.launch that contains the following.

<launch>
  <node name="usb_cam" pkg="usb_cam" type="usb_cam_node" output="screen" >
    <param name="video_device" value="/dev/video0" />
    <param name="image_width" value="320" />
    <param name="image_height" value="240" />
    <param name="pixel_format" value="mjpeg" />
    <param name="camera_frame_id" value="usb_cam" />
    <param name="io_method" value="mmap"/>
  </node>
  <node name="image_view" pkg="image_view" type="image_view" respawn="false" output="screen">
    <remap from="image" to="/usb_cam/image_raw"/>
  </node>
</launch>

Launching the file should display images from the camera.

roslaunch ~/ros/launch/usb_cam.launch


After that you can install the I Heart Robotics ROS Packages

cd ~/ros/stacks
git clone git://github.com/IHeartRobotics/iheart-ros-pkg.git
cd ~/ros/stacks/iheart-ros-pkg/ihr_demos/ihr_opencv
rosdep install ihr_opencv
rosmake ihr_opencv


The prerecorded video is stored in .bag files and can be viewed by running the launch file.

roslaunch ~/ros/stacks/iheart-ros-pkg/ihr_demos/ihr_demo_bags/launch/demo.launch




Next we will be showing how to convert the video from blue, green, red (BGR) to the hue, saturation, value (HSV) color-space.

Please post any problems with these instructions in the comments.