This ROS node will find the AprilTag 0 in the video stream and publish its pose in the TF tree

Dependencies:

• ROS Kinetic
• OpenCV3
• Visp3 and visp_bridge (sudo apt-get install ros-kinetic-vision-visp)
• libxml2

Installation: mkdir build && cd build && cmake .. && make install

Usage: rosrun agimus-vision tracker_box

There are some ros parameters used to configure the node:

• /imageTopic: Topic where the camera publish the RGB image (default: /camera/rgb/image_rect_color).
• /cameraInfoTopic: Topic where the camera publish the information about the camera settings (default: /camera/rgb/camera_info).
• /positionNode: Name of the TF node corresponding to the box position (default: box).
• /positionParentNode: Name of its parent node in the TF tree (default: rgbd_rbg_optical_frame).wn

Generate DAE from AprilTag png files

Moved to gerard bauzil package.

Debugging images on the robot

• We emulate a virtual display on the robot. On the robot, run the command in a terminal and leave it open.

Xvfb :1 -screen 0 800x600x16

• We start the tracker node within this display, by setting the environment variable DISPLAY to :1 for the tracker node executable.

export DISPLAY=:1

• Once the node is launched, you can ask for the debug to be published. There are two parameters for this: ~debugDisplay and ~publishDebugDisplay. Something similar to rosparam set /vision/tracker/debugDisplay true followed by rosparam set /vision/tracker/publishDebugDisplay true should enable the publication.

rosparam set /vision/tracker/publishDebugDisplay true
rosparam set /vision/tracker/debugDisplay true

• On another terminal, possibly on another computer, run:

rosrun image_view image_view image:=/vision/tracker/debug _image_transport:=compressed