Background of the plan
The deepening and implementation of government digital transformation and the application of new smart cities have driven the development of IoT public data platforms; The development of Internet of Things, big data, and artificial intelligence technology has provided strong technical support and driving force for the development of IoT public service platforms; At present, the data situation of smart city business is complex: there are multiple platforms, formats, and sources, making it difficult to utilize their value; There are multiple data resource management platforms, lacking overall planning and collaboration;
Content of the plan
Establish a dual base of technology and operation, create intelligent and data dual engines, attract and support more government departments and research and development institutions to develop richer smart business applications based on big data, provide modern governance macro situation analysis and decision-making support for urban leadership, provide smart governance business support for government departments, provide accurate and efficient livelihood services for the public, and continue to support the construction of new smart cities.
Advantages of the plan
1.Implement unified management and scheduling of algorithms, computing power, and tasks, including video resources, computing resources, storage resources, intelligent management and scheduling, and other content. Maximize utilization efficiency.
2. Intelligent resource allocation: Supports pre allocation of algorithms and computing power, and can configure different numbers of chips for different algorithms.
3. Analysis Task Management: The task management module provides scheduling capabilities for backend intelligent analysis tasks such as image detection modeling and video structuring, providing data support for the intelligent business of video analysis systems.
4.The algorithm repository is the core, which manages and stores various algorithm packages. The algorithm repository is an open system, where, in addition to the algorithms already built-in, external third-party algorithm packages can also be imported, or algorithm packages built through algorithm training platforms. The richness and diversity of algorithms in the algorithm repository are key points for mining the value contained in videos.
