From Perception to Phenotyping: Drone-based Simultaneous Localization and Mapping for Precision Woody Crop Management

Agriculture is facing growing pressure to produce more food with fewer resources and less environmental impact. My PhD research explores how drones can help farmers better understand orchards and vineyards. Instead of only taking pictures from above, drones can use cameras and artificial intelligence to estimate where they are, build 3D maps, and recognize important crop features such as trees and fruits.
This research studies how reliable these drone-based mapping methods are in real farming environments, where changing light, shadows, leaves, branches, and repeated tree rows make sensing difficult. It also develops methods to reconstruct fruit shapes and support crop monitoring in orchards, including challenging systems such as agrivoltaic orchards with solar panels.
The significance of this work is that it brings drone technology closer to practical precision agriculture, helping future farming systems monitor crops more accurately, reduce labour, and support more sustainable orchard management.