Empowering Farmers: Real-Time Coconut Tree Disease Detection and Management Using YOLOv8 and Machine Learning Technique

Coconut farming occupies a critical place in tropical farming economy, but its development is retarded by several plant diseases of coconut that may decline the yield and quality of the crop. Most conventional methods of disease detection are sluggish, prone to human failures and not scalable. This is an advanced real time coconut tree disease detection system that uses YOLOv8 as a tool and explainable AI (XAI). The system has ability to process input from uploaded images, live webcam feeds and RTSP streams accurately to detect and classify multiple coconut tree diseases. To improve explainability, EigenCAM is integrated to visually show the affected regions and disease names/confidence scores directly over the images. A farmer friendly Streamlit interface allows the farmers make informed decisions. Compared to other models, it is efficient for real time applications for an average inference time of 20 milliseconds per image with almost 95% accuracy. We have shown high detection precision, increased trust with explainability, and user engagement through, which are a scalable and practical disease management solution for coconut plantations in the initial deployments.