Introduction

This is a very simple ROS package that provides a single node aimed at loading camera calibration info from disk and publishing it on a camera_info topic. Nearly all significant functionality from this node is provided by the camera_info_manager_py package. Normally, a camera driver incorporates the functionality of camera_info_manager_py or camera_info_manager directly into the driver. This is the preferred and default mode of operation.

However, in some situations this particular node may still prove useful. Here are a few example situations where one could use this node.

Bagged Data

Let’s say you bagged raw image data from a camera, but you forgot to bag the camera calibration data as well. One option would be to repair the bag file using one of the available rosbag APIs to insert the missing camera data, but perhaps an easier option would be to run this package to publish the corresponding camera_info topic at the same time as playing back the bag file.

Comparing Calibrations

Let’s say you are using image_geometry to perform some real-world measurements with a camera, and you would like to compare the relative accuracy of two different camera calibrations. You could easily run a single camera driver that loads up one of the calibrations, and then use topic_tools relay to relay the raw camera data to a new topic or namespace. Then you could use this package to load the second camera calibration on a different topic.

Manufacturer Provided Calibrations

Some cameras that I’ve encountered provide calibration info out-of-the-box, and overriding that calibration info is not always easy. This package would provide a relatively simple way of loading an additional camera calibration topic for you to use.

Usage

The package only provides a single node camera_info_manager_node that essentially only instantiates an instance of the camera_info_manager.CameraInfoManager class, and then publishes the loaded camera data on the camera_info topic at a set frequency. As a result, nearly all of the documentation for the relevant parameters, topics, and services is found in the camera_info_manager_py package. For convenience, I’ll provide a brief description below:

Published Topics

camera_info

Contains camera calibration info as a standard sensor_msgs/CameraInfo message

Services

set_camera_info

A standard sensor_msgs/SetCameraInfo service that allows you to modify data being published on camera_info. Calling this service will not modify a loaded YAML file.

Parameters

~freq

(default: 10) A float that says the frequency that you’d like to publish the camera_info topic at.

~cname

(default: ‘camera’) A string representing the camera name. This should match the name in whichever YAML calibration file is loaded. If you don’t specify the ~url parameter, then this can also be used to automatically find the correct YAML file in ~/.ros/camera_info/

~url

(default: ”) A string representing a Uniform Resource Locator for the YAML file that you want to load. Examples from the documentation:

			   - `file:///full/path/to/local/file.yaml`
			   - `package://camera_info_manager_py/tests/test_calibration.yaml`
			   - `package://ros_package_name/calibrations/camera3.yaml`
    

Note you can also use ${NAME} and ${ROS_HOME} substitution variables.

~namespace

(default: ”) A string representing a string that you’d like to prepend to the set_camera_info service. Often it may be preferable to use the ns roslaunch attribute or the __ns key with rosrun.