On-farm evaluation of UAV-based aerial imagery for season-long weed monitoring under contrasting management and pedoclimatic conditions in wheat

Timely availability of weed infestation maps is a key prerequisite for the implementation site-specific management practices. Low-altitude aerial imagery obtained from unmanned vehicles (UAVs) has shown significant potential detection in crops. However, most studies focused on wide-spaced row crops such as maize and sunflower evaluated proposed methods single or few well-characterized experimental sites at specific points time representing limited range application scenarios. This study feasibility on-farm wheat fields characterized by narrow spacing, throughout early late developmental stages using UAV ground-based high-resolution imagery. Image data was nine sites, wide pedoclimatic conditions. These can be seen representative sample scenarios that would encountered practice. A high within- across-site well temporal variation observed levels population species composition, highlighting need spatially temporally resolved mapping. Image-based classification vegetation objects crop plants achieved with an accuracy 0.88 0.72 images UAV-based captured altitude 10 m, respectively. The pixel-wise, vegetation-index-based estimation during vegetative varied strongly across sites. Our results highlight critical importance ground resolution object-based image analysis growth robust are applicable suggested future research aiming rapid should focus development systems. Yet, monitoring stands currently available equipment offers reducing pressure control measures context rotations.