With the rapid development of Internet of Things (IoT) technology, smart agriculture has become a new direction in the development of modern agriculture. This study aims to enhance the utilization of agricultural resources and strengthen the intelligent and scientific management of crop growth. Given the low efficiency, high cost, and insufficient monitoring accuracy of the current traditional agricultural monitoring methods, this study analyzes the key technologies in the domestic and international agricultural IoT field based on the current situation of China’s agriculture. A smart agricultural monitoring system is proposed that combines convolutional neural network (CNN), IoT, and blockchain technology. The system addresses issues such as low efficiency, excessive cost, and inadequate monitoring precision present in traditional agricultural information systems. Meanwhile, it adopts a modular design to enhance scalability and employs a multitasking scheduling mechanism similar to an operating system to ensure high stability. This study also proposes a blockchain-based IoT data platform and designs simulated experiments to verify the system’s performance. The dataset used in the study includes photographic materials of corn from germination to maturity across six growth stages, covering an area of approximately 15 square meters, with images captured from 9 a.m. to 6 p.m. each day. Moreover, it selects images from the corn planting period, encompassing a vast amount of image data across six growth stages, of which about 20% of data is used for the test set. The results show that the recommended model has different recognition rates for corn crops at diverse growth stages, with an overall recognition error rate, weighted averaged, ranging only from 0.18% to 3.53%. In addition, the model can predict factors affecting crop growth with high precision. The blockchain data storage model still demonstrates excellent system performance under high transaction throughput, meeting the performance requirements of IoT systems. This study fills several research gaps in intelligent agricultural management information systems and provides practical guidance and insights for the application effects of “blockchain + deep learning” technology in sustainable rural development.
Qi et al. (Sat,) studied this question.