aruco_detector

ROS package to detect ArUco markers in images and estimate their pose.

Prerequisites

OpenCV

Eigen3

e.g., in Ubuntu

sudo apt install libopencv-dev libeigen3-dev

Installation

Clone within a working ROS workspace $WS

cd $WS/src
git clone https://github.com/xenvre/aruco_detector

Build

cd $WS
catkin build

Usage

Once a camera is plugged in your rig and its node is running, simply use the roslaunch file provided:

roslaunch aruco_board_detect aruco_board_detect.launch [show_debug_image:=true]

Some roslaunch parameters you might find useful:

Parameter	Description
`camera_info_topic`	The node will subscribe to this topic to source camera parameters
`camera_image_topic`	The node will subscribe to this topic to source input images
`show_debug_image`	Shows the output image in a OpenCV window
`detection_period`	Time between marker detections (seconds)
`markers_config_file`	Config file for marker settings

Please check the launch file for defaults. Specifically, the file cfg/markers_config.yaml is used as default configuration file.

Published topics

Topic	Explanation
`/aruco_board_detector/marker_pose`	The marker poses (stamped with the camera reference frame)
`/aruco_board_detector/debug_image`	The output image, i.e. the input image with markers drawn on it

The node will also broadcast several tf frames named marker_<id> where <id> is the ID of the marker.

Marker generator

The following command shows how to generate an ArUco marker. In the example, we consider a 4.0 x 4.0 cm (-s 4.0) marker with ID = 0 (-i 0) from the dictonary DICT_5X5_50 (-t DICT_5X5_50).

python `rospack find aruco_detector`/scripts/generate_aruco.py -o marker.png -s 4.0 -i 0 -t DICT_5X5_50

For futher options, run the above command with --help: