Solution Architecture

The overall technical architecture consists of an AI Vision Box, network cameras, a management backend, a mobile app, an NVR, and other components, providing a complete closed-loop solution from front-end intelligent monitoring to back-end data processing.

Solution Value

01/Reduced Labor Costs

① AI algorithms perform real-time intelligent detection of surveillance content with high accuracy, effectively reducing errors and omissions, eliminating the need for multiple manual guards.

② Automatic alarms are triggered upon detection of abnormal states, with real-time upload of alarm information, effectively reducing the workload of inspection personnel.

③ Automatic analysis of surveillance content, self-tagging, and cloud storage of effective surveillance information effectively reduce the workload for surveillance queries.

02/Enhanced Venue Safety

① High-accuracy AI facial recognition algorithms provide real-time monitoring and early warning for strangers and blacklisted individuals, ensuring the safety of personnel in the venue.

② Intelligent analysis of personnel behavior provides real-time alarms for misconduct, dangerous actions, falls, and injuries, preventing hazards from occurring immediately.

③ Real-time detection of conditions within the scene, providing remote early warnings for incidents such as fires and equipment damage.

03/Data Storage and Security

① Supports cloud storage or autonomous local area network storage, auditable and controllable, with no risk of data leakage, ensuring data security.

② Supports upgrades from local or cloud sources, including full package and incremental package upgrade management for software applications and algorithm components.

③ Domestically developed and independently controllable, utilizing domestic chips and proprietary algorithms, ensuring reliability and controllability.

04/Convenient Device Deployment

① The AI Vision Box is plug-and-play, with built-in algorithms compatible with existing security systems, independently controllable hardware, reducing deployment costs.

② Compatible with various general-purpose network cameras, supporting standard video streaming protocols (e.g., RTSP/GB28181) and high-definition resolution.

③ Local technical team support, including device installation and deployment training services, enabling rapid go-live for convenient device deployment.

Smart Hardware

AIBOX AI Vision Box

Product Specifications

Max Video Analysis Channels 6 channels 1080P

Video Hardware Decoding Capability 8 channels HD video (1080P)

CPU RK3588 Octa-core, main frequency up to 2.4GHz, Linux/Android 12.0/Domestic OS, 6 TOPS computing power

GPU Quad-core Mali-G610 MP4 high-performance GPU

RAM/Storage Standard 8GB/Standard 64GB

Display Interfaces 2 HDMI outputs (up to 8K resolution), 1 DP output, 1 HDMI input

External Ports Type-C1, USB3.01, RS2322, GPIO2, RS4851 (optional), CAN1 (optional), LINE IN/OUT*2

Internal Ports USB2.01, Mini PCI-e1 (optional), M.2 SSD (optional), SATA (optional)

Network Configuration Wireless Network: Built-in WiFi6; Bluetooth 5.0 Wired Network: Supports 2 channels 10/100/1000M adaptive Ethernet; Mobile Network: Supports PCI-E interface 4G internet access (optional)

Operating Temperature 0℃~60℃

Box Dimensions 166164.252mm

Application Scenarios

Smart surveillance has been adapted with a rich set of AI algorithms, compatible with over thirty algorithms including fire and smoke detection, work uniform recognition, illegal intrusion, abnormal loitering, smoking detection, facial recognition, crowd gathering detection, fall detection, distress signal detection, leaving post detection, phone call detection, safety helmet detection, and climbing detection.