Computer Vision-Based Autonomous Navigation for UAV Vineyard Row Following

This study investigates the implementation of a Robotics operating system (ROS) enabled Unmanned Aerial Vehicle (UAV) simulation for vineyard row following. The primary motivation stems from the imprecision of GPS positioning in agricultural areas, where high accuracy is crucial for tasks such as crop irrigation. The proposed approach utilizes a computer vision algorithm to accurately determine the vineyard row position and, therefore, adjust the UAV navigation. This algorithm, combined with ROS, is tested on a PX4-enabled UAV. The findings indicate that our method offers significantly improved precision over the blind positioning obtained through a Graphical user interface (GUI) based ground control system which does not take into account the position of the vineyard row with respect to the UAV but only the end position manually selected by the user.