Authors:
S. Rubin Bose, J. Angelin Jeba, N. Christy Evangeline, R. Regin, M. Mohammad Sameer Ali, S. J. Vimal Aravintha

Addresses:
School of Computer Science and Engineering, SRM Institute of Science and Technology, Ramapuram, Chennai, Tamil Nadu, India. Department of Electronics and Communication Engineering, S.A. Engineering College, Chennai, Tamil Nadu, India. Department of Electronics and Instrumentation Engineering, Madras Institute of Technology, Chennai, Tamil Nadu, India. Department of Research and Development, Dhaanish Ahmed College of Engineering, Chennai, Tamil Nadu, India. Department of Data Science, Indiana University Bloomington, Bloomington, Indiana, United States of America.

Abstract:

This study presents AgroGraphNet, an innovative framework for predicting and analysing crop disease spread using GNNs integrated with satellite imagery. The proposed system models farm as graph nodes and establishes edges based on spatial proximity and environmental similarity, enabling effective modelling of disease propagation across agricultural regions. By leveraging multi-source data, including Sentinel-2 satellite imagery, weather parameters, and soil characteristics, AgroGraphNet learns complex spatial–temporal dependencies that influence disease dynamics. The system employs GNN architectures such as GCN, GraphSAGE, and GAT to capture inter-farm relationships and predict potential disease outbreaks. Experimental evaluations demonstrate that AgroGraphNet provides accurate, interpretable, and scalable predictions suitable for diverse agricultural conditions. The framework includes a CLI package that enables researchers and practitioners to preprocess data, train models, and visualise prediction outcomes efficiently. This research highlights the potential of combining remote sensing and graph-based learning to improve early disease detection and response in agriculture. The integration of spatial modelling, satellite analytics, and GNNs supports precision farming, enhances policymakers' decision-making, and contributes to sustainable agricultural management. Future work will focus on expanding data sources, optimising temporal modelling, and integrating real-time disease-monitoring capabilities.

Keywords: Command-Line Interface (CLI); Remote Sensing; Precision Agriculture; Artificial Intelligence (AI); Geospatial Modelling; Sustainable Farming; Graph Neural Networks (GNNs); Graph Convolutional Networks (GCN); Graph Attention Networks (GAT).

Received on: 16/12/2024, Revised on: 23/02/2025, Accepted on: 15/04/2025, Published on: 07/12/2025

DOI: 10.69888/FTSIN.2025.000549

FMDB Transactions on Sustainable Intelligent Networks, 2025 Vol. 2 No. 4, Pages: 176-189

• Views : 90
• Downloads : 21

Download PDF