#60 · Primary category: Robotics & Embodied AI Simulation

Autonomous-Ai-drone-scripts

ai ardupilot autonomous-quadcoptor computer-vision gps lidar

State of the art autonomous navigation scripts using Ai, Computer Vision, Lidar and GPS to control an arducopter based quad copter.

Project last updated:04/07/22

GitHub Stars

622

Forks

170

Contributors

8

License

Other

Why we included this project

Most drone AI repos stop at a demo video, but this one includes code that has actually flown. The person-following system pairs a MobileNet object detector on the live camera feed with a TF Luna LiDAR for range, then steers yaw and roll through PID controllers in guided mode. Everything runs on a Jetson Nano, which is a useful reference if you want to keep inference and control loops on the drone itself rather than offloading to a ground station. You also get a data plotter for PID debug output and a Keras-to-TensorRT path for speeding up models. One honest caveat: the end-to-end point-to-point pilot has moved into a commercial product, so the open part is mainly the follow-person stack and tooling. Treat it as experimental and test in the ArduPilot SITL simulator before real flights.

Articles for this project

No articles for this project yet.

To suggest a topic or contribute an article, contact us.

Related projects in this category