Mapping Indicator Species of Segetal Flora for Result-Based Payments in Arable Land Using UAV Imagery and Deep Learning
A recent study published in Ecological Indicators (December 2024) explores an innovative approach to monitoring indicator species of segetal flora in arable fields using UAV imagery and deep learning. The research, conducted by Caterina Barrasso and colleagues, addresses the challenges in conserving biodiversity on farmland, particularly for species affected by intensive agriculture in Europe. By utilizing UAVs with RGB cameras and the YOLO deep learning model, the study demonstrates a cost-effective method for identifying key species across winter barley fields under varied management intensities.
The team successfully mapped half of the observed species, achieving species-specific detection rates between 49-100%. Critical to the process was a ground sampling distance (GSD) of 1.22 mm and the role of plant height, which significantly improved detection accuracy. Additionally, spatial co-occurrence with UAV-detectable species and canopy height heterogeneity offered valuable insights for locating less visible species.
The findings highlight UAV-based deep learning as a promising, scalable solution for supporting result-based payments (RBPs) and conserving biodiversity in agricultural landscapes. This research provides a foundation for smarter, more accessible monitoring practices for segetal flora, supporting future biodiversity efforts across Europe.
For more details, read the full article here: https://doi.org/10.1016/j.ecolind.2024.112780.