Bienvenue À

XCORE AI Smart Box

Accuracy, Reliability, Versatilit

XCORE AI Smatr box Artificial Intelligence Workstation is an intelligent video analysis processing terminal device from HFSECURITY that supports running deep learning neural network algorithms.The internal integrated high computing power AI processing module, up to 28.8Tops of super computing power. High-performance configuration of the main control platform, with 16 1080P HD video decoding capabilities, integrated with a variety of video signal processing units, with lightweight, high performance, low power consumption, easy access, etc., widely used in various fields such as face recognition, intelligent security, behavior analysis

XCORE SMART AI Box (5)

Artificial Intelligence Workstation

XCore Intelligent edge computing server Supports simultaneous processing of 16 1080P@30fps H.264/H.265 live streams or stored streams

XCORE SMART AI Box (5)

Rich Interface

XCORE Intelligent Analysis AI BOX Support 28.8Tops deep learning operation, using a combination of front-end image preprocessing + back-end algorithm operation, supporting low-bit quantization of CNN, RCNN
and other neural network structure of deep learning dedicated gas pedal, support lowbit quantization of AlexNet, VGG, ResNet and other classification neural network,
support low-bit quantization of Faster R-CNN, SSD, FPN and other target detection
neural network.

XCORE SMART AI Box (5)

Artificial Intelligence Workstation

Support dual Gigabit network ports, expandable 4G/WIFI wireless network mode

XCORE Series

xcore smart ai box
xcore smart ai bos

Cours Particuliers

HFSECURITY fingerprint scanner offers an unparalleled level of security and convenience for a wide range of applications. With its advanced features and compatibility with multiple operating systems, it is a versatile and essential tool for any organization or individual who values security and privacy.

FACILITÉ D'UTILISATION

QUALITÉ D'IMAGE CERTIFIÉE

GRANDE VITESSE

MULTIFONCTIONNEL

xcore smart box

XCORE Product Features

  1. Supports simultaneous processing of 16 1080P@30fps H.264/H.265 live streams or stored streams.
  2. Support 28.8Tops deep learning operation, using a combination of front-end image preprocessing + back-end algorithm operation, supporting low-bit quantization of CNN, RCNN and other neural network structure of deep learning dedicated gas pedal, support lowbit quantization of AlexNet, VGG, ResNet and other classification neural network, support low-bit quantization of Faster R-CNN, SSD, FPN and other target detection neural network.
  3. Support face detection, tracking, recognition, etc.
  4. Support dual Gigabit network ports, expandable 4G/WIFI wireless network mode.
  5. Low power consumption design, combined with shell cooling, can increase the cooling fan active
    cooling, product stability and reliability.
  6. Fully embedded structure, highly integrated, can support desktop placement or wall mounting,
    very easy to install and maintain.

Device access

  1.  Xcore AI Video analysis box High-definition network cameras that comply with O NVIF , RTSP , and GB/T28181 standards
  2. Xcore Edge Computing Server the face capture machine that conforms to the GA / T 1400 protocol
  3. Support H.265 (does not support web preview) / H.2 64 video standard, up to 8 million pixel IPC access

Smart application

  1. The total library supports up to 300,000 face pictures, and supports 64 libraries
  2. Multi-class algorithm mixing: face mode, mixed mode (face-human binding + video structured)
  3. Face recognition: support full load maximum 4/8/1 6 video streams
  4. Video structure: support full load maximum 4/8/1 6 video streams
  5. Perimeter alert: support full load maximum 4/8/1 6 video streams
  6. Behavioral alert: support a maximum of 4/8/10 video streams with full load
  7. Item alert: support full load maximum 4/8/16 video streams
  8. People counting: support full load maximum 4/8 video streams
  9. Fire alert (industry extended algorithm package): supports a maximum of 4/8/16 video streams at full load
  10. Production safety supervision (industry extended algorithm package): supports a maximum of 4/8/16 video streams at full load
  11. Bright Kitchen Bright Stove (Industry Extended Algorithm Package): Supports a maximum of 4/8 video streams at full load
  12. Support face capture, face recognition comparison alarm, stranger recognition alarm, etc.
  13. Supports recognition of face attributes: gender, age, hat, glasses, mask, etc.
  14. Supports face, human body, motor vehicle, non-motor vehicle, license plate capture and face-human binding, human-non-motor vehicle binding
  15. Support the analysis of human body attributes such as the color of the upper and lower clothing, the style of the upper and lower clothing, the status of the backpack, whether to wear a helmet, etc.
  16. Support motor vehicle classification, vehicle attribute analysis such as color, brand, driving direction and non-motor vehicle classification
  17. Support license plate recognition and license plate-motor vehicle binding relationship
  18. Support perimeter alert: no parking of vehicles , departure of vehicles, loitering of personnel, over-the-wall detection, intrusion, border crossing
  19. Supportive behavior alert: running, falling, smoking , making phone calls, looking at mobile phones, sleeping on duty, leaving duty, overcrowded on duty , less on duty , gathering of people, scuffling
  20. Support item alert: Debris piled up
  21. Support people statistics: regional people statistics , entrance and exit people statistics
  22. Support industry algorithm expansion: open kitchen and bright stove , fire warning , production safety supervision
  23. Support bright kitchen and bright stove: mouse detection , garbage bin uncovered detection 
  24. Support fire alert: detection of smoke , flame , and fire facilities
  25. Support production safety supervision: safety helmets , overalls , safety belts, reflective clothing , flames, smoke, fire protection facilities detection
XCORE SMART AI Box (5)

System characteristics

  1. Support MegConnect protocol to access Hongtu, Pangu and other software platforms
  2. Cloud map or third-party view library platform through GA / T 1400 standard protocol
  3. Provide rich RESTful API protocol interface, which is convenient for third-party platforms to do docking development

Application Scenario

  1. Provides the capabilities of face recognition, all-target structured analysis, intelligent warning , and people counting , and cooperates with the upper platform software to realize the closed loop of face capture recognition, video structuring, intelligent warning , and passenger flow counting
    scenarios.
  2. Applied to places that require face recognition, such as entrances and exits of smart communities, entrances and exits of office buildings, key personnel list control and identification alarm, stranger identification, etc.
  3. Applied to structured places that require face-human-machine non-plate capture and license plate recognition, such as the control of people and vehicles in smart parks, and license plate recognition in smart communities.
  4. It provides industry algorithm expansion capabilities such as production safety supervision, bright kitchen, fire protection, etc., and cooperates with the industry platform to realize closed-loop scenarios such as safe production, hygienic kitchen, and fire protection
System Parameters
Main ProcessorHigh performance embedded microprocessor
Operating SystemEmbedded Linux
Device Access
Video Stream InputVideo resolution:1920 x 1080 (2 million ), 2560 x 1440 ( 4 million), 3840 x 2160 (8 million )
Video Decoding TypeH .264 / H .2 65
Face Capture MachineMaximum support for 8 /16/ 32-channel capture machine access (the capture machine must support the GA / T 1400 protocol)
Smart Prameters
Human Face RecognitionCapture rate 99%
Capture false capture rate< 1%
Whitelist identification pass rate> 99.5 % _
Recognition false recognition rate: < 0.5%
The largest base library: 300,000 people portrait library
Human Face RecognitionCapture rate 9 5 %
Capture false capture rate< 1%
Motor VehicleCapture rate 90 %
Capture false capture rate< 1%
Non-motor VehicleCapture rate 9 5 %
Capture false capture rate< 1%
License PlateCapture rate 9 5 %
Capture false capture rate< 1%
Recognition rate 9 5 %
Interface parameters
Network Interface2, 10 0M/1000 M adaptive Ethernet , RJ 45 interface
Alarm Input/Output Interface4 switches
Audio Input/Output1 way
Front/ RearUSB InterfaceFront 1 USB2.0 , 1 USB 3.0; Rear 2 USB 2.0
R S4852 way
Reset Button/Power LED ( PWR)1
Running Indicator Light (RUN)1 pc, two-color (red+green)
System functions
Running Indicator Light (RUN)Passage of employees at entrances and exits; key personnel list control and identification alarm, stranger identification, etc.
Video StructuredCapture, attribute analysis, license plate recognition of faces, human bodies, motor vehicles, non-motor vehicles, license plates, etc.
Algorithm Warehouse Managementalgorithm package upload via web, API , Meg Connect protocol, support algorithm package installation , uninstallation, deletion
Network ProtocolTCP/UDP/HTTP/MULTICAST/DHCP/FTP/NTP/HTTPS/RTSP /GAT1400 etc
Dual Network PortAre supported : 1.multiple addresssettings 2.load balancing 3.Active/standby mode
Log QueryInquiry, search and display of face-human-machine non-card capture information
Environmental equirements
Température de fonctionnement-30 ~+ 70
Température de stockage-30 ~+ 70
Relative Humidity1 0% ~ 90%RH, non-condensing
Autre
Power SupplyDC12V±10%, 2A
Structuremetal case
Dimensions (L, D, H)229 × 193 × 49 (mm)
Poids<2kg

Présentation De L'Entreprise

Chongqing Huifan Technology Co., Ltd. est une société chinoise qui fournit des solutions biométriques et des technologies associées. En tant que fournisseur de solutions biométriques, ils proposent une gamme de produits et services liés à la vérification d'identité et au contrôle d'accès.

Nous proposons des produits, des services et des technologies. , il est courant que les entreprises du secteur des solutions biométriques proposent diverses technologies biométriques, telles que la reconnaissance des empreintes digitales, la reconnaissance de l'iris, la reconnaissance faciale, la reconnaissance vocale, etc.

HFSECURITY propose des solutions pour des secteurs ou des applications spécifiques, telles que le suivi du temps et des présences, l'authentification des paiements, l'identification personnelle, etc.

Si vous avez des questions spécifiques sur Chongqing Huifan Technology Co., Ltd., vous pouvez visiter notre site officiel ou nous contacter directement pour plus d'informations.

CEO

Mike

CSO

Sienna

CMO

Vicky

CTO

Adam

Télécharger la brochure du scanner d'empreintes digitales

xcore smart ai box
xcore smart ai bos

Customized

If you want to know more about the algorithm, or if you have your own demand scenario, fill out the form below to get a customized solution!

Packaging customization

    votre nom
    Nom de l'entreprise

    Facebook
    https://twitter.com/HF_Security
    LinkedIn

    Application Scenarios XCORE