Deep Learning-Driven Soccer Drone for Agricultural Health Monitoring System

Abstract

Agricultural health monitoring systems are critical for guaranteeing food security and increasing crop yields by detecting plant diseases in a timely and accurate manner. Traditional monitoring approaches frequently rely on manual inspections, which are time-consuming, labour-intensive, and susceptible to human mistake, thereby delaying essential solutions. In contrast, modern monitoring systems leverage advanced technologies such as drones, sensors, and deep learning algorithms to continuously track crop health in real-time, enabling precise and targeted interventions. The main goal of this work is to integrating advanced drone technology with deep learning algorithms for real-time monitoring and disease classification in paddy and coconut trees. The drone, assembled using high-performance components, it facilitates efficient and high-resolution imaging under real-world agricultural conditions. The collected visual data is subsequently processed using deep learning techniques to identify and classify diseases affecting paddy and coconut trees. In particular, Residual Network (ResNet) architecture was employed for disease prediction and its performance was benchmarked against a conventional CNN model. Experimental results demonstrate that ResNet outperforms the CNN model, achieving higher accuracy and robustness in disease detection.