<![CDATA[This research focuses on the development of a grapevine disease identification system based on leaf imagery using the YOLO (You Only Look Once) algorithm. Disease identification in grapevines presents a significant challenge in agriculture due to its impact on productivity and harvest quality. The primary issues in this identification include dataset limitations, variability in leaf imagery, and similarities in symptoms between diseases, which can reduce detection accuracy. Additionally, the computational demands of YOLO models pose challenges for IoT devices with limited processing capabilities, while the lack of integration with environmental data further complicates the system. This study offers a solution by developing a YOLO model trained on a dataset of 480 leaf image samples. The dataset includes various grapevine leaf diseases under different lighting conditions, image angles, and diverse backgrounds. This dataset was used to train the model to detect and identify specific diseases in grapevine leaves. The system was subsequently tested with unseen leaf images to evaluate accuracy, precision, sensitivity, and processing speed. The results demonstrate that the YOLO model achieved an average accuracy of 86.8% and a precision of 73.6%, with fast processing times suitable for real-time application. However, challenges such as dependency on dataset quality and hardware requirements remain key concerns. The system proved capable of recognizing and classifying several types of grapevine leaf diseases effectively, making it a valuable tool for early disease detection. With broader implementation and adjustments to local conditions, this technology is expected to reduce production losses and enhance harvest quality.]]>
Copyrights © 2025
