Explore the docs »

View Demo · Report Bug · Request Feature

The goal of our “Duck Vision” project was the integration of the OAK-D device from OpenCV’s AI Kit into the Duckietown platform. We think that the integration of OAK-D in the Duckietown infrastructure can boost autonomous car software development and research, as well as expose students who are working in the Duckietown framework to spatial AI and stereo vision.

1. Run on your terminal dts --set-version daffy-oakd.
2. Switch on you Duckiebot and disable the official camera: dts duckiebot camera --stop.
3. Mount the OAK-D on the robot using the camera mount.

To integrate the basics of the OAK-D on a Duckiebot (tested with DB21M), you need to:

1. Clone the OAKD-Tutorials repository: git clone [email protected]:viciopoli01/oakd-tutorials.git
2. Run and use the notebooks we developed, navigate to oakd-tutorials/oakd and run: dts oakd notebooks. At this point a Jupyter lab starts and you will be redirected on a web page. Login with the password quackquack.
3. You can find the folders 01-basic-integration, 02-oakd-calibration, 02-oakd-calibration, 03-semantic-segmentation. Each of these folders contains a notebook, follow the instructions and have fun!

To create advanced pipeline with the OAK-D on you Duckiebot you can:

1. Use the OAKD-template to create your DepthAI application.
2. Create a ROS package that uses the DepthAI Library and DTROS to have a fll integration in the Duckietown ecosystem.
3. To build and run your application on the Duckiebot use the dts devel run/build commands. Refer to the Duckietown docs for more information.

For easily mount the OAK-D on your Duckiebot we designed a camera support you can print. The files are in the folder camera_mount.

We assume you are familiar with the Duckietown environment and you have already you computer setup to work with a Duckiebot as showed here.

Contributions are what make the open source community such an amazing place to be learn, inspire, and create. Any contributions you make are greatly appreciated.

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

Vincenzo Polizzi - @linkedin - [email protected]

Trevor Phillips - @linkedin - [email protected]

Project Link: DuckVision

Thanks to:

• Duckietown, for letting us using the lab.
• Microsoft Azure and intel for sponsoring the OpenCV AI Competition 2021.
• othneildrew for the README template.