The integration of computer vision (CV) and generative artificial intelligence (GenAI) into smart agriculture has revolutionised traditional farming practices by enabling real-time monitoring, automation, and data-driven decision-making. This review systematically examines the applications of CV in key agricultural domains, such as crop health monitoring, precision farming, harvesting automation, and livestock management, while highlighting the transformative role of GenAI in addressing data scarcity and enhancing model robustness. Advanced techniques, including convolutional neural networks (CNNs), YOLO variants, and transformer-based architectures, are analysed for their effectiveness in tasks like pest detection, fruit maturity classification, and field management. The survey reveals that generative models, such as generative adversarial networks (GANs) and diffusion models, significantly improve dataset diversity and model generalisation, particularly in low-resource scenarios. However, challenges persist, including environmental variability, edge deployment limitations, and the need for interpretable systems. Emerging trends, such as vision–language models and federated learning, offer promising avenues for future research. The study concludes that the synergy of CV and GenAI holds immense potential for advancing smart agriculture, though scalable, adaptive, and trustworthy solutions remain critical for widespread adoption. This comprehensive analysis provides valuable insights for researchers and practitioners aiming to harness AI-driven innovations in agricultural ecosystems.

The deep integration of artificial intelligence (AI) and Internet of Things (IoT) technologies has promoted the application of Artificial Intelligence of Things (AIoT) in smart agriculture, driving substantial transformations in agricultural data management systems and facilitating sustainable agricultural development. This paper reviews the key AIoT data management technologies employed in smart agriculture, systematically exploring the technical applications and practical implementations across four architectural layers. These layers (i.e. perception layer, transport layer, data platform layer, and application layer) correspond to agricultural information sensing, data transmission, storage processing, and intelligent decision-making, respectively. Research indicates that, through the deployment of diverse sensing devices, low-latency short-range communications, wide-area wireless networks, cloud-edge collaborative computing, blockchain-based data storage, and AI-driven agricultural decision support systems, the data management framework of smart agriculture is undergoing continuous and intelligent enhancement. AIoT exhibits significant potential in enhancing agricultural productivity, optimizing resource allocation, and improving environmental adaptability. However, significant challenges remain, particularly in protecting agricultural data privacy and achieving real-time interoperability across heterogeneous platforms. The paper concludes by outlining future research directions, offering valuable insights for researchers in the field of smart agriculture.

The introduction of a consortium blockchain-based agricultural machinery scheduling system will help improve the transparency and efficiency of the data flow within the sector. Currently, the traditional agricultural machinery centralized scheduling systems suffer when there is a failure of the single point control system, and it also comes with high cost managing with little transparency, not leaving out the wastage of resources. This paper proposes a consortium blockchain-based agricultural machinery scheduling system for solving the problems of single point of failure, high-cost, low transparency, and waste of resources. The consortium blockchain-based system eliminates the central server in the traditional way, optimizes the matching function and scheduling algorithm in the smart contract, and improves the scheduling efficiency. The data in the system can be traced, which increases transparency and improves the efficiency of decision-making in the process of scheduling. In addition, this system adopts a crowdsourcing scheduling mode, making full use of idle agricultural machinery in the society, which can effectively solve the problem of resource waste. Then, the proposed system implements authentication access mechanisms, and allows only authorized users into the system. It includes transactions based on digital currency and eliminates third-party platform to charge service fees. Moreover, participating organizations have the opportunity to obtain benefits and reduce transaction costs. Finally, the upper layers supervision improves the efficiency and security of consensus algorithm, allows supervisors to block users with malicious motives, and always ensures system security.

Large-scale and heterogeneity of the Internet of Things (IoT) sensing networks introduce a big challenge to device connectivity. There exist some isolated nodes in randomly deployed IoT sensing networks running on a tree-typed topology due to limitations of some network parameters, which reduces network connectivity. In this paper, a connectivity enhancement scheme for the sensing networks of the IoT is proposed based on link transformation. Under constraints of network depth and the number of child nodes, we boost capability of an in-network node to connect more isolated nodes by reducing its or ancestors' depth. Furthermore, three-level node shifting is utilized to take full advantage of network locality, thus highly improving ability of a potential parent node to accept connection request of an isolated node. Finally, when failing to reduce depth of a node and disabling to shift out a child node of a parent node, the scheme exploits node swapping to improve present link status, thus enabling further some isolated nodes to join into the sensing networks. Our simulation results show that the proposed scheme can raise proportion of joined nodes and effectively enhance connectivity of the sensing networks in the IoT.

The advent of sensor-cloud technology alleviates the limitations of traditional wireless sensor networks (WSNs) in terms of energy, storage, and computing, which has tremendous potential in various agricultural internet of things (IoT) applications. In the sensor-cloud environment, virtual sensor provisioning is an essential task. It chooses physical sensors to create virtual sensors in response to the users’ requests. Considering the capricious meteorological environment of the outdoors, this paper presents an measurements similarity-based virtual-sensor provisioning scheme by taking advantage of machine learning in data analysis. First, to distinguish the changing trends, we classified all the physical sensors into several categories using historical data. Then, the k-means clustering algorithm was exploited for each class to cluster the physical sensors with high similarity. Finally, one representative physical sensor from each cluster was selected to create the corresponding virtual sensors. The experimental results show the reformation of our scheme with respect to energy efficiency, network lifetime, and data accuracy compared with the benchmark schemes.