shawnbmccarthy/viam-ros-camera-wrapper

Viam ROS Camera Wrapper

This has been tested with ROS 1 Noetic, Melodic testing is underway

ROS Configuration

Our robot was configured with a standard USB camera on both the Melodic and Noetic releases. The video_stream_opencv package was used.

If roscore is not running, make sure it is running in one window

roscore

Note: for our local testing we only start ROS when testing

In another window start our camera publisher node

roslaunch video_stream_opencv camera.launch

If there are multiple cameras we need to set the appropriate video stream provider argument.

List the camera topics

rostopic list 
/camera/camera_info
/camera/camera_stream/parameter_descriptions
/camera/camera_stream/parameter_updates
/camera/image_raw
/camera/image_raw/compressed
/camera/image_raw/compressed/parameter_descriptions
/camera/image_raw/compressed/parameter_updates
/camera/image_raw/compressedDepth
/camera/image_raw/compressedDepth/parameter_descriptions
/camera/image_raw/compressedDepth/parameter_updates
/camera/image_raw/theora
/camera/image_raw/theora/parameter_descriptions
/camera/image_raw/theora/parameter_updates
/rosout
/rosout_agg

Noetic

Tested & working

Melodic

Currently still testing with ROS Melodic

To consider

Possible issues

When installing viam-sdk on jetson running ROS 1 Melodic, it will be required to download and compile Python3.9

Setup Viam Remote

git clone https://github.com/shawnbmccarthy/viam-ros-camera-wrapper
cd viam-components-camera-wrapper
./setup_venv.sh

This will install our python virtual environment and install the required packages

Configure Robots remote resources

The viam server component interface allows us to create custom components which interface with the viam-server using the viam SDK's. Our example makes use of the Python SDK to create custom sensors.

1. log into app.viam.com

2. access the robot you configured, and go to the modules tab

3. give the module a name, for this example lets use ros-module as the name, next set the executable to the run_module.sh (using the full path):

4. Now that we have created our module, we need to add the components, select Components tab

5. this module contains two components, a camera and a sensor. To add the camera enter:

   1. enter general name for the camera component (ros_camera_demo or anything you prefer)
   2. Select camera from the type drop down
   3. enter viamlabs:ros:camera as the type
   4. select create component
   5. in the attributes window enter {"topic": "ROS_CAMERA_TOPIC"}, where ROS_CAMERA_TOPIC is the topic that produces images
   6. select save

6. to add the sensor component

   1. enter general name for the sensor component (ros_sensor_demo or anything you prefer)
   2. Select sensor from the type drop down
   3. enter viamlabs:ros:sensor as the type
   4. select create component
   5. in the attributes window enter {"topic": "ROS_SENSOR_TOPIC"}, where ROS_SENSOR_TOPIC is the topic that produces images
   6. select save

Now go back to app.viam.com and select the Control tab of your robot and select Sensors, here we can select Get All Readings to see the output

And go to the camera to view the image

To see how the code works view: camera_remote.py and ros_camera.py

Further activities

Testing and updates to add more topics

contact

For any questions, please email me at: shawn@viam.com

References

1. viam
2. viam cloud
3. viam docs
4. viam python sdk
5. sdk as server
6. subclassing components
7. drive a rover