The AI industry involves many scenarios, with huge market space and incremental space. Among them, the more mature areas are mainly concentrated in public security, transportation, communities, parks and schools. The AI video surveillance system is the core of smart public security. The architecture of the public security AI video surveillance system is mainly divided into three layers. Terminals collect information and real-time monitoring; the edge layer performs facial recognition comparisons on part of the video streams and image streams accessed by the front end. Realize structured attribute analysis, recognition and storage; the analysis layer is to build a dynamic comparison and recognition system and a static portrait system, and connect the collected data to the system for analysis and management.
A public security department in Xinjiang has purchased face recognition AR glasses, which are mainly used in airport personnel activity security checks and personal identification comparison, etc., to improve law enforcement efficiency.

Product Features
Smart police AR glasses have functions of portrait recognition and license plate recognition. The product has five major features in terms of appearance and functional performance: split design, anti-sunlight interference, multi-thread processing, wake-up-free voice operation, and high recognition accuracy.
Folding temples, compact and portable
The mirror legs of the fuselage are made of aviation-grade magnesium-aluminum alloy, with an ultra-light weight of 90 grams, ensuring a light and senseless wearing experience. Patented design high-precision shaft, through 20,000 times of high-strength bending experiments, it can be easily folded and stored, compact and portable.
Ultra-clear display, wide field of view
Using industry-leading array optical waveguide technology, the transparency, contrast, and color accuracy are all higher than the industry average. The 400:1 ultra-high contrast ratio brings an unparalleled clear and lifelike display effect, and the 40° ultra-wide viewing angle brings the visual experience equivalent to an 85-inch HD TV 3 meters away.
Full voice, hands-free
Equipped with a self-developed full-voice operating system that fully integrates AI voice and AR. The whole operation supports wake-up-free speech recognition, and special optimizations have been made in noise reduction and directional recognition, making the response speed as fast as 0.5 seconds and the recognition accuracy rate as high as 98%.
NPU blessing, fast and accurate identification
Equipped with NPU neural network processing chip, using leading image recognition algorithms, it has strong performance in data processing, image acquisition, and face recognition. The speed of face recognition under the support of NPU is as fast as 0.6 seconds, and it can recognize up to 10 faces at the same time, which fully meets the needs of the security field.
Large battery, 8 hours long battery life
Equipped with an external DOCK, built-in 10000mAh fast charge large battery, 3 times the battery capacity of a smartphone. In the working environment, the battery life is up to 8 hours. At the same time, DOCK supports 4G, GPS and other expansion modules to meet the needs of customers in different industries.

extensions
Voice assistant
Customizable language control system, supports developers to customize Chinese online/offline voice recognition commands, and provides the most intuitive and convenient user interaction
Face recognition
Supports simultaneous recognition of 20 dynamic faces within a distance of 4 meters, the corresponding time is less than 200 milliseconds, and the accuracy rate is as high as 99.8%, providing developers with world-class face recognition capabilities
Object recognition
Supports any object recognition model customization and 360-degree recognition, and the response time is less than 200 milliseconds in a human environment, providing developers with world-class object recognition capabilities
Remote collaboration
Support 1080P video calls, provide augmented reality communication, precise visual annotation, private cloud deployment and other capabilities, and provide safe and reliable cross-platform real-time collaboration services for enterprises and public institutions
Head control
Provides a head control interactive development kit based on a built-in gyroscope. Based on this, developers can quickly complete the development of functions such as head control-based information browsing, unlimited screen, and information follow.
Vision kit
Provides a visual development kit optimized for the characteristics of the interactive fish reality of glasses, virtual and reality alignment tools, etc., so that developers without AR development experience can quickly complete AR applications with good experience.

AR smart glasses solutions with recognition systems can be widely used in public security, traffic police, urban management, fire protection, and security inspections for major events. In addition to face recognition, AR smart glasses can also compare human IDs and obtain results in real time. In case of emergencies, the first-person visual live broadcast can be quickly performed.