joselusl / aruco_eye

aruco_eye

This is the repository of the aruco_eye metapackage. It is a ROS wrapper of the ArUco library (ArUco: a minimal library for Augmented Reality applications based on OpenCV http://www.uco.es/investiga/grupos/ava/node/26), adding extra functionalities very useful for Robotics.

Table of contents

• Disclaimer
• Repository information
  • Packages
  • Libraries and executables generated by this repository
  • Branches
• Repository requirements and dependencies
  • System Requirements and Compatibility
  • Dependencies
    • Additional System Dependencies
    • ROS Dependencies
    • Extra ROS Dependencies
• Installation Instructions
  • Pre-requirements
  • Installation steps
• Running Instructions
  • aruco_eye_ros_detector
    • ROS Launcher
    • ROS Topics
    • ROS Parameters
    • ROS Services
    • Additional Configurations in files
  • aruco_eye_ros_display
    • ROS Launcher
    • ROS Topics
    • ROS Parameters
    • ROS Services
    • Additional Configurations in files
• Documentation
• License
• Contributors
  • Authors / Maintainers / Past contributors
  • Contact information
  • New contributors
• Acknowledgments

Disclaimer

You are using this software at your own risk. The authors decline any responsibility for personal injuries and/or property damage.

Repository information

Packages

This repository is formed by a five ROS packages:

• aruco_lib: original ArUco library (see http://sourceforge.net/projects/aruco/ and http://www.uco.es/investiga/grupos/ava/node/26 ) with catkin elements (CMakeLists.txt and package.xml)
• aruco_eye_core: C++ (and ROS-independent) wrapper of the ArUco library with additional functionalities.
• aruco_eye_ros: ROS-dependent wrapper of the ArUco library with additional functionalities.

And one ROS metapackage:

• aruco_eye

Libraries and executables generated by this repository

This repository generates:

• Libraries: several
• Executables: two, aruco_eye_ros_detector and aruco_eye_ros_display

Repository Branches

This repository has the following branches:

• Master: Catkin version of the package.

Repository requirements and dependencies

System Requirements and Compatibility

This package is running under:

• Ubuntu: 14.04 and above
• ROS: Indigo and above with Catkin

This package has been tested under:

• Ubuntu: 14.04
• ROS: Indigo and Jade with Catkin

Dependencies

Additional System Dependencies

This metapackage requires:

• OpenCV 2.0 and above
• Boosts

ROS Dependencies

This metapackage depends on the following ROS packages:

• roscpp: http://wiki.ros.org/roscpp
• message_generation: http://wiki.ros.org/message_generation
• message_runtime: http://wiki.ros.org/message_runtime
• std_msgs: http://wiki.ros.org/std_msgs
• sensor_msgs: http://wiki.ros.org/sensor_msgs
• geometry_msgs: http://wiki.ros.org/geometry_msgs
• image_transport: http://wiki.ros.org/image_transport
• message_filters: http://wiki.ros.org/message_filters
• cv_bridge: http://wiki.ros.org/cv_bridge
• tf2: http://wiki.ros.org/tf2
• tf2_geometry_msgs: http://wiki.ros.org/tf2_geometry_msgs
• tf2_ros: http://wiki.ros.org/tf2_ros

Extra ROS Dependencies

This metapackage depends on the following extra ROS packages:

• pugixml: https://github.com/joselusl/pugixml
• perception_msgs: https://github.com/joselusl/perception_msgs
• robot_component_srvs: https://github.com/joselusl/robot_component_srvs

Installation Instructions

Pre-requirements

Install all the system and ROS dependencies (for example, using rosdep install, or apt-get, or directly from source).

Installation steps

1. Create your ROS catkin workspace as usual.
2. Inside your workspace, download the extra ROS package pugixml: git clone https://github.com/joselusl/pugixml ./
3. Inside your workspace, download the aruco_eye repository: git clone https://github.com/joselusl/aruco_eye ./
4. Compile your ROS catkin workspace as usual.

Running Instructions

The package aruco_eye_ros included in this repository generates two ROS nodes that can be executed:

aruco_eye_ros_detector

This node processes the given images searching for ArUco visual markers, publishing them in a ROS topic.

ROS Launcher

The package aruco_eye_ros includes a launcher called aruco_eye_ros_detector.launch.

ROS Topics

Subscribed topics:

• camera/image_raw [sensor_msgs/Image]: The images where the ArUco visual markers are going to be detected.
• camera/camera_info [sensor_msgs/CameraInfo]: The calibration info of the camera.

Published topics:

• aruco_eye/aruco_observation [perception_msgs/MarkerList]: The list of the detected visual markers with all its information.
• tf [tf2_msgs/TFMessage]: The tf information of the detected and reconstructed visual markers.

ROS Parameters

Configurable ROS parameters:

• aruco_list_file [type=string, default=$(find aruco_eye_ros)/config/arUcoList.xml]: the path to the ArUco list configuration file. See below for more information.
• camera_calibration_file [type=string, default=]: the path to the camera calibration file. If nothing is used here, it uses the information given by the camera/camera_info topic.
• aruco_detector_frame_name [type=string, default=aruco_detector]: the name of the aruco detector frame.
• aruco_marker_child_base_name [type=string, default=aruco_marker_]: the base name for the visual markers frame.
• image_topic_name [type=string, default=camera/image_raw]: the topic name of the images.
• camera_info_topic_name [type=string, default=camera/camera_info]: the topic name of the camera calibration.
• aruco_list_topic_name [type=string, default=aruco_eye/aruco_observation]: the topic name for the list of detected visual markers.

ROS Services

None.

Additional Configurations in files

There are two configuration files. An example can be find in aruco_eye_ros/config:

• camera.yml: camera configuration file in YAML format. File type generated by OpenCV. For more information, please refer to: http://www.uco.es/investiga/grupos/ava/node/26

• arUcoList.xml: ArUco list configuration file in XML format. It has the information of the ArUco visual markers that want to be detected. Please open the example file in aruco_eye_core/config/arUcoList.xml for basic information. Please, be sure that the format is correct, otherwise the node will not work.

aruco_eye_ros_display

This node allows the visualization of the detected ArUco visual markers by the node aruco_eye_ros_detector.

ROS Launcher

The package aruco_eye_ros includes a launcher called aruco_eye_ros_display.launch.

ROS Topics

Subscribed topics:

• camera/image_raw [sensor_msgs/Image]: The images where the ArUco visual markers are going to be detected.
• camera/camera_info [sensor_msgs/CameraInfo]: The calibration info of the camera.
• aruco_eye/aruco_observation [perception_msgs/MarkerList]: The list of the detected visual markers with all its information.

Published topics:

• aruco_eye/aruco_observation_image/image_raw [sensor_msgs/Image]: The image with the drawed ArUco Visual Markers.

ROS Parameters

Configurable ROS Parameters:

• display_output_image_window_name [type=string, default=arUco_display_window]: the name of the OpenCV local window to display the visual markers.
• flag_display_output_image [type=boolean, default=True]: flag to enable or disable the display of the drawed markers using an OpenCV local window
• display_output_image_service_name [type=string, default=aruco_eye/enable_display_output_image]: service name to enable or disable the display of the drawed markers using an OpenCV local window.
• camera_calibration_file [type=string, default=]: See above.
• camera_info_topic_name [type=string, default=camera/camera_info]: See above.
• image_topic_name [type=string, default=camera/image_raw]: See above.
• aruco_list_topic_name [type=string, default=aruco_eye/aruco_observation]: See above.
• output_image_topic_name [type=string, default=aruco_eye/aruco_observation_image/image_raw]: topic name of the image with the drawed ArUco Visual Markers.

ROS Services

Services served:

• aruco_eye/enable_display_output_image [robot_component_srvs/SetBool]: to enable or disable the display of the drawed markers using an OpenCV local window.

Additional Configurations

There is one configuration file. An example can be find in aruco_eye_ros/config:

• camera.yml: camera configuration file in YAML format. File type generated by OpenCV. For more information, please refer to: http://www.uco.es/investiga/grupos/ava/node/26

Documentation

For more information related to the ArUco library or camera calibration, please refer to: http://www.uco.es/investiga/grupos/ava/node/26.

It exists another similar package called aruco (see http://wiki.ros.org/aruco and https://github.com/pal-robotics/aruco_ros) with the following limitations compared to aruco_eye:

• It is just a wrapper of the ArUco library with few examples and it does not implement a complete solution. The metapackage aruco_eye implements a complete solution for robotics.
• The detection and the drawing is done by the same node, what can be a problem in robotics, for example, when processing onboard and having a ground station for visualization. In aruco_eye is solved by splitting the detection and 3D reconstruction node (aruco_eye_ros_detector node) and the drawind and visualization (aruco_eye_ros_display node),
• It is hard to implement a big number of visual markers with different sizes. In aruco_eye is solved using an XML config file.

This repository is the continuation of the work done by Jose Luis Sanchez-Lopez to use the ArUco library in ROS, initied in January 2013. The first version was splitted in the following repositories (already outdated and deprecated) and its dependencies:

• https://bitbucket.org/joselusl/lib_aruco
• https://bitbucket.org/joselusl/arucoeye
• https://bitbucket.org/joselusl/dronearucoeye
• https://bitbucket.org/joselusl/dronearucoeyerosmodule

Unfortunately this repository has no more extra documentation.

If this repository gets the interest of the community, we will continue adding information depending on the questions made by its users.

License

All distributed software are subject to the 3-clause BSD license. See the LICENSE file.

Please, refer to http://www.uco.es/investiga/grupos/ava/node/26 to check the license of the ArUco library (BSD).

Contributors

Authors / Maintainers / Past contributors

List of people that have contributed:

• Jose Luis Sanchez-Lopez ([email protected]): Main Author, Maintainaince, Debugging, Testing and Documentation.

Contact information

Questions, suggestions, requests, ... whatever. Feel free to contact me (Jose Luis Sanchez-Lopez - [email protected]).

New contributors

You are very welcome to contribute to this repository by opening a pull request via Github.

Acknowledgments

Special thanks to:

• Computer Vision Group (CVG): www.vision4uav.eu
• Centre for Automation and Robotics (CAR): http://www.car.upm-csic.es/
• Consejo Superior de Investigaciones Cientificas (CSIC): http://www.csic.es
• Universidad Politecnica de Madrid (UPM): http://www.upm.es

aruco_eye's Issues

Error with opencv version

If you have opencv 2.x and want to run this code

you probably get an error....

Solution::

use this code:
so make out of the 3 a 5
cause:

https://stackoverflow.com/questions/22702555/does-opencv-have-a-constant-with-its-version-number

#if  CV_VERSION_MAJOR >= 5
    cv::Ptr<cv::ml::SVM>  _model;
#else
    cv::Ptr<CvSVM>  _model  ;
#endif
public:
    SVMMarkers(){
        // static variables from SVMMarkers. Need to be here to avoid linking errors
        _minFeatureValue = -1;
        _maxFeatureValue = 1;
        _patchSize = 10;
        _dictSize = -1;
        _rotateMarkers = true;

    }


    int getBestInputSize(){return _patchSize;}
    bool load(string path)throw (cv::Exception){
    if (path.empty())return false;


#if  CV_VERSION_MAJOR >= 5
    _model= cv::ml::StatModel::load<cv::ml::SVM>(path);
     _patchSize=sqrt(_model->getSupportVectors().size().width);

#else
    _model=new CvSVM;
    _model->load(path.c_str());
    _patchSize=sqrt(_model->get_var_count());

#endif

How to set the aruco coordinate system?

Hey there the previous coordinate system looks like this for me:

But I would like to set somewhere in the code a coordinate system flag such that the z axis of the aruco marker points out of the image.... (and the x,y axis are also turned to be a right hand system)

(currently z is pointing down)

Best
Markus