Multimodal deep learning framework for LoRaWAN based crop health monitoring
TL;DR
This study proposes AgriPredict–LoRaWAN Network (APLNet), an Internet of Things (IoT)-based agricultural monitoring and disease classification framework designed for cucumber farming that supports timely decision-making, improves productivity, reduces crop losses, and promotes sustainable cucumber cultivation.
Abstract
Cucumber is a widely cultivated vegetable crop worldwide. However, environmental stress, pest infestations, and diseases can significantly reduce crop quality and yield. Early disease detection and continuous crop monitoring are essential for improving productivity and minimising agricultural losses. This study proposes AgriPredict–LoRaWAN Network (APLNet), an Internet of Things (IoT)-based agricultural monitoring and disease classification framework designed for cucumber farming. The system integrates LoRaWAN (Long Range Wide Area Network) enabled sensors to continuously collect critical soil parameters, including moisture, temperature, pH, and electrical conductivity. In addition, leaf image analysis is employed to identify healthy and diseased cucumber plants. The collected soil and leaf data are processed and analysed using the proposed APLNet model, and the model’s performance is evaluated using standard classification metrics, including accuracy, precision, recall, and F1-score. Experimental results demonstrate that APLNet achieves superior classification performance compared with established deep learning models, including EfficientNet-B0 to B7, ResNet50, and InceptionV3. The proposed model achieves 98% classification accuracy in distinguishing healthy from unhealthy cucumber leaves. Furthermore, the proposed system is deployed as a web-based decision-support AgriPredict platform that enables farmers with real-time crop monitoring and management. It supports timely decision-making, improves productivity, reduces crop losses, and promotes sustainable cucumber cultivation.