AI-Based IoT System for Soil Nutrition and Plant Disease Detection in Smart Agriculture

Abstract

The increasing global population and changing climate conditions have created a strong demand for sustainable and intelligent agricultural practices. Smart agriculture, powered by Artificial Intelligence (AI) and the Internet of Things (IoT), provides effective solutions for real-time soil nutrition monitoring and early plant disease detection. IoT-based agricultural systems continuously collect environmental and soil data through sensors, enabling accurate monitoring of moisture, temperature, nutrient levels, and crop conditions. Deep learning techniques such as Convolutional Neural Networks (CNNs), Residual Networks (ResNet), and Vision Transformers (ViT) have significantly improved plant disease detection and classification by automatically extracting complex features from plant images and sensor data. These models outperform traditional machine learning methods in accuracy, scalability, and automation. The integration of IoT and deep learning supports intelligent decision-making, precision farming, and improved crop productivity. Advanced hybrid architectures combining residual learning, coordinate attention, and transformer-based attention mechanisms further enhance feature extraction by capturing both local spatial details and global contextual dependencies. These approaches improve disease identification accuracy and soil analysis performance. This paper presents a comprehensive review of AI techniques for IoT-based soil nutrition monitoring and plant disease detection systems, highlighting recent developments, hybrid deep learning models, optimization techniques, and challenges such as computational complexity, heterogeneous data integration, and deployment limitations in smart agriculture environments.