Drone-Based Weed And Disease Detection In Agricultural Fields To Maximize Crop Health Using a Yolov8 Approach

The advancement of drone technology has opened up new opportunities for precision agriculture, enabling efficient monitoring and management of agricultural fields. This study introduces a new method that utilizes the YOLOv8 model to detect and classify weeds and diseases in agrarian fields using aerial images captured by drones. The model was trained on a batch size of 32 for 40 to 100 epochs, and the hyperparameters were fine-tuned to get optimal performance. This resulted in a precision of 87% and a recall of 80% for weed detection and 60% precision and 43% recall for disease detection. The results indicate the potential of drone-based weed and disease detection as a valuable tool for farmers and agronomists to gather accurate, timely, and comprehensive data about their fields, facilitating informed decision-making and precise management practices. By leveraging the advantages offered by drones, farmers have the opportunity to enhance the efficiency of resource allocation, increase crop productivity, ensure food security, minimize their ecological footprint, and actively contribute to long-term sustainability farming methods that preserve natural resources for future generations.