UxV-Swarm based Automated Fruits Disease Instance Segmentation for Smart Farming using Deep Learning Model

Smart farming is a revolutionary method in agriculture that utilizes advanced technologies to improve crop monitoring and disease segmentation. This research study investigates the usage of unmanned vehicles (UxV), such as Unmanned Aerial Vehicles (UAVs) and Unmanned Ground Vehicles (UGVs), in a swarm-based system to segment fruit diseases in smart farming automatically. The proposed technique utilizes deep learning (DL), notably YOLOv8 instance segmentation, to segment diseases in various fruit datasets accurately and efficiently. The suggested approach improves the efficiency and precision of disease segmentation and enables immediate monitoring and action. The UxVs collaborate using a swarm strategy, resulting in a scalable solution for extensive agricultural activities. By using DL models, notably YOLOv8, one may achieve the level of accuracy necessary for complex tasks like fruit disease segmentation. The study integrates a wide range of datasets that include many types of fruits and their linked diseases, such as Strawberry, Tomato, and numerous leaf diseases. The DL model is trained on a merged dataset, resulting in a remarkable overall accuracy of 95%. Furthermore, the model's strong performance in disease segmentation across diverse crops is evident from the individual dataset assessments, which indicate high accuracy compared with other existing studies.