In this epic project, you’ll learn how to precision park your Jetson Nano Smart Rover autonomously with OpenCV
The standard Jetson Nano Smart Rover already has the camera you’ll need in the kit. Then all you have to do is print out Aruco 72 (basically a QR Code) and paste it on a hard substance like wood or even cardboard.
What is precision parking you ask? Good question. Well the concept comes from drones, where conventional drones are only about as accurate as plus or minus about 30 feet. Uh, that’s not accurate. The term precision landing means landing a drone on a target with centimeter level accuracy. This same 30 foot radius problem is the same with rovers, because GPS just isn’t accurate enough. Precision Parking means parking a rover on a waypoint with centimeter level accuracy, and performed autonomously.
You can also do some fun “Follow-Me” missions with your rover with the exact same script used for precision parking. So if you want to take your rover for a walk, just carry the aruco marker, run the script and your rover WILL CHASE YOU!
Below is a summary of everything covered in this project:
- Dependency downloads
- Calibrating camera for Accurate OpenCV readings
- Code overview of Precision Parking Script
- Demo of Precision Parking
- Demo of Follow Me
This project is based on our Rover kits. You’ll need that for this project. After receiving the rover, there are no further required hardware add-on purchases.
Members can access the code for this Voice Assistant project in our private GitHub Repo by clicking the button above. If you get a 404 error, it means we haven’t added you to the private repo. Request access with the button to the right.
If you are a Drone Dojo member, you will have access to every private repo for all of the projects. But first, you’ll need to request access so that we can authenticate your account. Click the button above to complete this one-time process.