Towards Robotization of Foraging Wild Fruits Under Canopy - A Multi-camera Drone-Borne Berry Mapping

The rapid technological advancements are allowing not only the automatization of work, but often also its facilitation through human-robot collaboration. This topic is investigated in the FEROX project, which addresses the underexplored domain of improving the process of wild berry harvesting in Northern European forests with robotics and AI. This paper investigates the integration of multi-camera drone technology for under-canopy mapping in the context of wild berry location mapping. Our proposed methodology lays a groundwork for utilizing AI methods to provide georeferenced maps of berries’ locations in forest areas, inherently characterized by an unreliable GNSS signal. We carry out initial tests in a forest in eastern Finland with a custom hexacopter, proving the suitability of our approach for retrieving a geographical position of detected fruits with the tested sensor configuration, enabling further processing to supply foragers with wild fruit yield heat maps on a per-species basis.