Autonomous Vision-based UAV Navigation

From RSN
Jump to: navigation, search

We study the problem of vision-based autonomous navigation of an autonomous vehicle in an orchard row.

Our goal is to enable vision-based Unmanned Aerial Vehicle (UAV) fully autonomous navigation in orchard-like environments for the purposes of data collection without relying on GPS.

We have developed custom controllers and vision based obstacle detection and avoidance algorithms.

For the vision component, we design a general purpose obstacle avoidance algorithm using a frontal stereo camera. For the controller component, we design a controller that uses a downward facing camera and provides reliable performance even on the presence of strong wind disturbances.

Fully autonomous Orchard Navigation under strong wind

UAV view while performing Fully autonomous Orchard Navigation

Fully autonomous Orchard Navigation - a difficult case

Fully Autonomous Yield Coverage in an Apple Orchard : Count the tags

Reconstructed UAV trajectory using only collected yield data

This project is funded by USDA NIFA MIN-98-G02.

Related Publications

1N. Stefas, H. Bayram, V. Isler
Vision-Based UAV Navigation in Orchards
In 5th IFAC Conference on Sensing, Control and Automation Technologies for Agriculture, 2016.