City Visual Intelligence Engine is built on the Alibaba Cloud distributed computing and storage platform. It uses advanced video, image, graphics processing, and deep learning algorithms to create city-level artificial intelligence models. The engine analyzes, indexes, and mines data from relevant scenarios to empower various industries, such as transportation, municipal governance, commerce, industrial parks, power and energy, and healthcare and education. City Visual Intelligence Engine integrates and connects visual resources from across a city. It uses images and videos to help the city think, make decisions, and operate, which improves the overall operational efficiency of the city.
Prerequisites
Before you read this topic, you must understand the following concepts:
Features
Data ingestion: Ingests, transcodes, and shards video and image data. Supports real-time online and offline ingestion from camera sources, such as bullet and dome cameras. Adapts to various ingestion protocols, such as GB/T 28181 and the GA/T 1400 view library.
Visual search: Creates indexes for objects, times, and events in videos. Supports efficient searches of semi-structured or structured data, such as high-dimensional vectors and properties. Supports efficient indexing technologies, such as feature quantization and heterogeneous index acceleration.
Data computing: Uses proprietary, world-leading algorithms for computer vision detection, recognition, and segmentation. Supports real-time analysis of hundreds of thousands of video streams and offline video analysis with 1,000× acceleration. Also supports sequence, graph, and graphics computing.
Industry engines: Empowers various industry scenarios, such as transportation, municipal governance, commerce, industrial parks, power and energy, and healthcare and education. This improves the overall operational efficiency of the city.
Benefits
Flexible, open, and secure platform: Built on the Urban Intelligence platform, one of China's first national open innovation platforms for artificial intelligence. It supports open algorithm integration and application development, offers flexible customization of services and workflows, and uses comprehensive model encryption and dynamic authorization mechanisms.
Advanced and extensive proprietary algorithms: Uses proprietary, world-leading AI algorithms for computer vision and deep learning.
Complete solutions: Offers on-demand intelligent analysis and complete, large-scale visual computing solutions to address pain points in various industries.
Global and intuitive digital city perception: Projects real-world information into a digital space to create a virtual real world. This enables full spatiotemporal awareness, full-element filter interaction, and full-cycle iteration for smart cities through global sandbox simulations and real-time command.