Kinova Vision module package

Overview

This ROS 2 package provides helper methods and launch scripts to access the Kinova Vision module depth and color streams.

Installation (using colcon)

The following instructions are for ROS 2, tested on Humble on Ubuntu 22.02.

Building from Source

Dependencies

GStreamer packages

• gstreamer1.0-tools
• libgstreamer1.0-libav
• libgstreamer1.0-dev
• libgstreamer-plugins-base1.0-dev
• libgstreamer-plugins-good1.0-dev
• gstreamer1.0-plugins-good
• gstreamer1.0-plugins-base

sudo apt install gstreamer1.0-tools gstreamer1.0-libav libgstreamer1.0-dev libgstreamer-plugins-base1.0-dev libgstreamer-plugins-good1.0-dev gstreamer1.0-plugins-good gstreamer1.0-plugins-base

Building

To build from source, clone the latest version from this repository into your catkin workspace and compile the package.

1. Create a workspace

mkdir -p ~/colcon_ws/src
cd ~/colcon_ws/src/

2. Clone this git repo into ~/colcon_ws/src

git clone https://github.com/Kinovarobotics/ros2_kortex_vision.git

cd ~/colcon_ws/
rosdep install --from-paths . --ignore-src -r -y
colcon build

Usage

Start kinova_vision node

source ~/colcon_ws/install/setup.bash
ros2 launch kinova_vision kinova_vision.launch.py 

The launch file provides arguments for launching depth, color, or registered depth images, as well as overriding other parameters. For example, to only launch the color node,

source ~/colcon_ws/install/setup.bash
ros2 launch kinova_vision kinova_vision.launch.py launch_depth:=false

Additional information is available below.

Please note that establishing a connection between the computer and the camera may require several attempts, so please be patient. Sometimes you may need to restart both the robot and the connected computer to successfully establish the connection.

Start rviz to view both cameras

Start rviz using the following command:

rviz2 -d src/ros2_kortex_vision/launch/camera_streams.rviz

This allows the visualization of both the RGB and the depth streams as well as the RGB point cloud.

Specifying launch options

It's possible to override the default argument values when launching the kinova_vision node.

Arguments are set using the following syntax: <argument>:=<value>.

For instance, the default value of the device argument can be overridden to specify another IP address.

ros2 launch kinova_vision kinova_vision.launch.py device:=10.20.0.100

Additional information on arguments color_camera_info_url and depth_camera_info_url

These arguments specify the custom camera information file to use instead of the default camera information file.

The file is specified via a specific URL syntax, using either of these two formats:

package://<package_name>/relative/path/to/file

file:///absolute/path/to/file

For example:

ros2 launch kinova_vision kinova_vision.launch.py color_camera_info_url:=file:///home/user/custom_color_calib_1280x720.ini depth_camera_info_url:=file:///home/user/custom_depth_calib_480x270.ini

A custom camera information file is typically created from a default information file (refer to launch/calibration/default_*.ini). Then, one simply needs to adjust the proper matrices.

The following matrices need to be adjusted with the proper values for FX, FY, PPX, PPY:

camera matrix
FX 0.00000 PPX
0.00000 FY PPY
0.00000 0.00000 1.00000

projection
FX 0.00000 PPX 0.00000 
0.00000 FY PPY 0.00000 
0.00000 0.00000 1.00000 0.00000 

The values for FX, FY, PPX, PPY can be obtained via the Vision module API. They represent the focal length and the principal point offset in both the x and y coordinates.

Nodes

kinova_vision_color

This node publishes the raw stream and the meta information of the color camera.

Subscribed Topics

None

Published Topics

• /camera/color/camera_info (sensor_msgs/CameraInfo)

  Color camera calibration and meta information.

• /camera/color/image_raw (sensor_msgs/Image)

  Color camera raw image (RGB8 encoding).

kinova_vision_depth

This node publishes the raw stream and the meta information of the depth camera.

Subscribed Topics

None

Published Topics

• /camera/depth/camera_info (sensor_msgs/CameraInfo)

  Depth camera calibration and meta information

• /camera/depth/image_raw (sensor_msgs/Image)

  Depth camera raw image (millimeters - 16UC1 encoding).

camera_color_tf_publisher

This node publishes the static coordinate transformation between the color camera frame (camera_color_frame) and the camera link frame (camera_link).

Subscribed Topics

None

Published Topics

• /tf_static (tf2_msgs/TFMessage)

  Color camera frame static transformation

camera_depth_tf_publisher

This node publishes the static coordinate transformation between the depth camera frame (camera_depth_frame) and the camera link frame (camera_link).

Subscribed Topics

None

Published Topics

• /tf_static (tf2_msgs/TFMessage)

  Depth camera frame static transformation

camera_nodelet_manager

This node uses the image_proc package to create a nodelet graph, transforming raw data from the device driver into point clouds, rectified images, and other products suitable for processing and visualization. First, make sure to install the image_proc ROS2 package on your system using the following command:

sudo apt install ros-$ROS_DISTRO-depth-image-proc

Then, start the launch file with the depth_registration argument set to true

ros2 launch kinova_vision kinova_vision.launch.py depth_registration:=true

Subscribed Topics

• /camera/color/camera_info (sensor_msgs/CameraInfo)

  Color camera calibration and meta information

• /camera/color/image_raw (sensor_msgs/Image)

  Color camera raw image

• /camera/depth/camera_info (sensor_msgs/CameraInfo)

  Depth camera calibration and meta information

• /camera/depth/image_raw (sensor_msgs/Image)

  Depth camera raw image

Published Topics

• /camera/depth_registered/camera_info (sensor_msgs/CameraInfo)

  Color camera calibration and meta information

• /camera/depth_registered/image_rect (sensor_msgs/Image)

  Depth rectified image (millimeters - 16UC1 encoding)

• /camera/depth/color/points (sensor_msgs/PointCloud2)

  Depth camera point cloud data with color information (RGBD)