Efficient Powdery Mildew Disease Detection Through a YOLOv4 and DenseNet Hybrid Model

This paper examines the applicability of advanced deep learning techniques, YOLOv4 and DenseNet in detecting powdery mildew, a common fungal pathogen from crops that causes a decline in agricultural yield. The main aim is therefore to develop a consistent and credible detection system that will assist with early identification. The models used were trained and tested using a large dataset of pictures of diseased plants and they were more effective than previous approaches. Classifying tasks saw DenseNet being ranked high, while object detection was done efficiently by YOLOv4. The overall accuracy of measures was noted to be at 95%. 8 % and thus the integrated strategy had a much higher success rate as compared to the conventional approaches of marketing. The findings of this study shall show that deep learning has the potential to enhance the control of plant diseases and secure appropriate farming practices.