视觉端边云融合架构: 面向超级智慧城市群演进的关键技术

The super metropolitan cluster is rising with the acceleration of urbanization. Technologies such as the Internet of Things, artificial intelligence, and big data are rapidly permeating the construction of city infrastructure, making smart cities a promising new field of interdisciplinary research. However, existing smart city solutions are typically customized for individual cities and lack a comprehensive intelligent perception of city operation and multi-regional coordination capabilities in practical applications. This paper proposes a visual end-edge-cloud (E²C) fusion architecture for the collaborative development of smart city clusters. Firstly, the new quality productive forces promoting the information infrastructure of smart cities are distilled into five-dimensional features: computability, transmission-utility, data-utility, intelligence-utility, and vision-utility. By constructing a hierarchical visual end-edge-cloud architecture among cities and intercity connectivity, the proposed solution addresses data interworking bottlenecks and breaks down intercity information islands, enhancing the flexibility and reliability of comprehensive systems. Additionally, feature-reconstruction-based resource-sharing techniques, cross-domain collaborative optimization for large-scale groups, and evolving complex city networks are introduced. These innovations enable intelligent connectivity among city clusters, forming a city-network architecture characterized by interworking, commonality, and sharing. This architecture supports smart city clusters in performing cross-scenario, precise, and real-time dynamic computations using visual data. This paper aims to provide a comprehensive technical architecture and solution for the development of smart city clusters to support the coordinated evolution of super metropolitan clusters.