Detection of Weeds Growing in Alfalfa Using Convolutional Neural Networks

Alfalfa (Medicago sativa L.) is used as a high-nutrient feed for animals. Weeds are significant challenge that affects alfalfa production. Although weeds unevenly distributed, herbicides broadcast-applied in fields. In this research, object detection convolutional neural networks, including Faster R-CNN, VarifocalNet (VFNet), and You Only Look Once Version 3 (YOLOv3), were to indiscriminately detect all weed species (1-class) discriminately between broadleaves grasses (2-class). YOLOv3 outperformed other networks detecting grass weeds. The performances of using image classification (GoogLeNet VGGNet) (Faster R-CNN YOLOv3) compared. GoogLeNet VGGNet (F1 scores ≥ 0.98) ≤ 0.92). Classifying training various broadleaf did not improve the performance detection. was most effective network 0.99) tested growing alfalfa. Future research will integrate into machine vision subsystem smart sprayers site-specific herbicide applications.