Apple Leaf Disease Prediction Using modified YOLOv8 Algorithm

Apple orchards face formidable challenges in disease management, with black rot, apple scab, and cedar apple rust being prominent fungal adversaries. Dealing with such an apple orchard disease are no small feat. This study introduces a sophisticated methodology to accurately spot and classify these diseases using the YOLOv8 model. The YOLOv8 model is fine-tuned, tweaking specific parts to make it more responsive to the unique traits of black rot, apple scab, and cedar apple rust leveraging in the PlantVillage dataset. The YOLOv8 model represents a better evolution within the YOLO series, developed to increase the detection speed while maintaining a great level of accuracy. The architecture consists of numerous convolutional layers, which enable feature extraction from input images, resulting in pinpoint accuracy in disease identification, the model undergoes end-to-end training, and it acquires the ability to learn from input data, improving its predictive performance for leaf disease instances, boosting an impressive accuracy rate of 99.5 %, demonstrating its ability to pinpoint and categorize apple leaf diseases with exceptional precision. Additionally, the model showcases a strong recall rate of 97.43%, confirming its effectiveness in accurately identifying instances of the targeted diseases. The outstanding accuracy and recall rates highlight the potential of our improved YOLOv8 model as a crucial asset in monitoring and controlling fungal infections impacting apple crops using artificial intelligence.