Early and on-ground image-based detection of poppy (<i>Papaver rhoeas</i>) in wheat using YOLO architectures

Abstract Poppy (also common poppy or corn poppy; Papaver rhoeas L., PAPRH) is one of the most harmful weeds in winter cereals. Knowing precise and accurate location essential for developing effective site-specific weed management (SSWM) optimized herbicide use. Among available tools mapping, deep learning (DL) used its accuracy ability to work complex scenarios. Crops represent intricate situations detection, as crop residues, occlusion weeds, spectral similarities between seedlings are frequent. Timely discrimination needed, because postemergence herbicides just when crops at an early growth stage. This study addressed P. detection wheat ( Triticum spp.) by comparing performance six DL-based object-detection models focused on “You Only Look Once” (YOLO) architecture (v3 v5) using proximal RGB images train models. The were assessed open-source software, evaluation offered a range results quality recognition well computational capacity during inference process. Of all models, YOLOv5s performed best testing phase (75.3%, 76.2%, 77% F1-score, mean average precision, accuracy, respectively). These indicated that under real field conditions, strategies can identify stage, providing information SSWM.