The calibration package

Calibrate any sensor network which provides ododometry messages of common hypothesis (trackings).

Graph Calibration

Performs a calibration of given sensors by graph optimization provided from g2o.

Preliminaries for Installation

Testet with ROS kinetik and Ubuntu 16.04

• Build and install g2o: https://github.com/RainerKuemmerle/g2o#compilation
• Build and install g2opy: https://github.com/uoip/g2opy#installation
• Substitute cmake .. with cmake -DPYBIND11_PYTHON_VERSION=2.7 ..
• Substitute python setup.py install with sudo cp lib/g2o.so /usr/lib/python2.7/dist-packages/g2o.so

Demo

• Start playing the bags and visualization: roslaunch calibration calibration_demo.launch
• Run calibration: rosrun calibration calibrator.py
• Store the calibrated data roslaunch calibration calibration_dump.launch
• Optional: Load the calibrated data roslaunch calibration calibration_load.launch

CITrack calibration

• Start cameras: roslaunch demo start_cameras.launch
• Start tracking wrt. marker setup: roslaunch demo start_aruco3_calibration.launch
• Record tracks in bags (stores bags to /tmp/cam[1,2,3,4].bag): roslaunch demo calibration_record.launch
• Stop tracking
• Start playing the bags and visualization: roslaunch calibration calibration_citrack.launch
• Run calibration: rosrun calibration calibrator.py
• Store the calibrated data roslaunch calibration calibration_dump.launch
• Optional: Load the calibrated data roslaunch calibration calibration_load.launch

Examples

$ ./inverse_transform.py -x 100. -y 2 -z 0 -R 0 -P 0 -Y 0
-100.0000 -2.0000 0.0000 0.0000 0.0000 0.0000 1.0000

static_inverse_transform_publisher does the following in a generic way:

$ rosrun tf static_transform_publisher $(./inverse_transform.py -x 100. -y 2 -z 0 -R 0 -P 0 -Y 0) source_frame target_frame 100

Nodes and Scripts

one_shot_calib_node

Records a single hypothesis multiple times and performs Slerp on angles and arithmetic mean on translations. The result is printed out as transformation from the parent to the child frame of the odometry message.

Name	Type	Default	Description
topic	string	/odom	Topic to record the hypothesis
percepts	int	1	Number of records to make from the hypothesis

dynamic_transform_publisher

Like a common transform_publisher, but with dynamic reconfiguration of the transformation parameters.

Name	Type	Default	Description
tf	string		Transformation from parent to child frame. Can be "x y z R P Y" or "x y z qx qy qz qw"
parent_frame	string		Parent frame id
child_frame	string		Child frame id
rate	double	1	Publish rate in Hz

calibration.py

Performs a calibration of given sensors by graph optimization provided from g2o.

1. Load calibration.yaml (initial sensor pose and path to rosbag tracking data) via rosparam.
2. calibration.py calibrates the sensors via dynamic reconfiguration. Therefore, all sensor tf's need to be set up by dynamic_transform_publisher.

inverse_transform.py

Calculates the inverse transformation and prints it to std::out so that it can be used by tf static_transform_publisher

odom_to_tf

This node updates a transformation between two coordinate frames with a given ros::nav_msgs::Odometry.

Parameter

Name	Type	Default	Description
ros_listener_odom_topic	string	/topic	Listenertopic for the ros::nav_msgs::Odometry to update the tf.
parent_frame	string	/parent	Parent frame id
child_frame	string	/child	Child frame id
rostimenow	bool	false	If this is set to true, ros::now as headertimestamp will be used otherwhise the timestamp from the input ros::nav_msgs::Odometry.