Citrus fruit disease prediction using ML on a decentralized and secure smart platform

Agriculture plays a vital role in society, and in recent years, researchers have increasingly turned to the Internet of Things (IoT) to gather data and develop innovative approaches to disease prediction. Storing all the gathered data in a Blockchain network has revolutionized the agricultural field. By integrating Blockchain, IoT, machine learning (ML), and cloud computing, farmers can now predict diseases more accurately while optimizing the use of natural resources. This paper used ML approach to classify citrus fruit diseases. IoT devices are used in agriculture to take data. These data are being passed through the gateway devices for classification. Various deep learning (DL) models such as AlexNet, VGG19, Xception, ResNet101, MobileNetV2, and ShuffleNet are employed to extract valuable features from the collected dataset. These DL models play a crucial role in feature extraction. The extracted features are then utilized for classification purposes, where a support vector machine (SVM) is commonly applied for classification. Additionally, an efficient device authentication mechanism is implemented to ensure the security and integrity of the system. The experimental results show that the Xception with SVM performs better than other models. Furthermore, the data is kept on a cloud-based Blockchain network to ensure stability, availability, immutability, security, and privacy in a smart agricultural setting.