YOLOv8-Enabled Real-Time Crop Health Monitoring with Conversational Diagnosis and Geospatial Support

YOLOv8-Enabled Real-Time Crop Health Monitoring with Conversational Diagnosis and Geospatial Support

Jampana Venkata Arjun Varma1, Talla Sunil Kumar2, Challa Yogesh3, Gade Prathyusha 4, Salapakshi Sagar 5, Dr. Afroz Pasha6

1School of Computer Science Engineering, Presidency University

2School of Computer Science Engineering, Presidency University

3School of Computer Science Engineering, Presidency University

4School of Computer Science Engineering, Presidency University

5School of Computer Science Engineering, Presidency University

6School of Computer Science Engineering, Presidency University

---------------------------------------------------------------------***---------------------------------------------------------------------

Abstract—Agriculture is a cornerstone of global economies, supplying food, employment, and raw materials for numerous industries. Yet, one of the sector’s enduring challenges is crop disease, which can drastically reduce yields and threaten food security. Traditional approaches to identifying plant diseases rely on manual inspections and expert evaluations, which are often slow, costly, and vulnerable to human error. Without early diagnosis, diseases can spread uncontrollably, leading to major economic setbacks for farmers and decreased crop output. To overcome these issues, this project introduces an AI- powered system for detecting and diagnosing crop diseases. It combines advanced deep learning, natural language processing (NLP), and geospatial mapping technologies. At its core is YOLOv8 (You Only Look Once, version 8), a powerful Convolu- tional Neural Network (CNN) designed for real-time image-based detection. Trained on a robust, annotated dataset from Roboflow, the model accurately identifies a variety of diseases affecting key crops such as rice, wheat, and maize.

Beyond detection, the system includes an intelligent chatbot powered by Large Language Models (LLMs). This virtual as- sistant offers instant, tailored advice on diagnosis, treatment options, and preventive strategies. It provides farmers with user- friendly guidance in natural language, making it accessible even to those with limited technical knowledge. The chatbot serves as a virtual agricultural consultant, recommending effective pesticides, organic treatments, and disease management practices. A standout feature of this project is its geospatial mapping capability. By integrating OpenStreetMap’s Overpass API, the system helps farmers locate nearby agricultural supply stores after a disease is identified. This allows quick access to the necessary products like pesticides or fertilizers, helping farmers

respond promptly to disease outbreaks. Overall, the system presents a comprehensive AI-driven approach to crop disease management by combining image- based detection, interactive chatbot support, and location-based resource mapping. By reducing the dependency on manual inspection, enhancing decision-making, and streamlining access to agricultural inputs, it promotes more efficient, tech-enabled farming. This real-time, intelligent solution not only boosts productivity but also minimizes economic losses, paving the way for a more sustainable and resilient agricultural future.

Keywords: Plant Disease Detection, YOLOv8 CNN, Agricul- tural AI, Intelligent Chatbot, LLMs in Farming, Plant Pathology AI, GIS Mapping, OpenStreetMap API, Smart Farming, AI- Enhanced Agriculture.