Hardware | Intel CPU: i7, Memory: DDR4 32GB RAM, HDD: 2.5" 500GB SSD |
OS | Linux |
Maximum Number of Faces in Database | 50,000 |
Network | 10/100Mb/1Gb Network |
Maximum Number of Devices Connecting | (Build-in 1CH) Up to: 6 |
Power Input | 100-240V AC |
Power Consumption | Max. 90W |
Dimensions | 240 x 210 x 57 mm (LWH) |
Weight | Weight: 1.3 kg net and 2.1 kg gross |
Saftity Certifications | CE, FCC, UL, CB, BSMI |
System |
Architecture | Three Tier Server-client system:- Web Server
- Application Server
- Database Server
|
Supported OS | Ubuntu Server 20.04 |
Supported Protocols | HTTP & HTTPS |
Supported Hypervisors | VMWare, Hyper-V, Virtual Box |
Container Type | Docker |
Database Type | MongoDB |
System Security | - Single device login
- Renewable session token
- Hardened HTTPS
|
Management Interface |
Interface | Using Web client |
Supported Browsers | - Google Chrome
- Mozilla Firefox
|
Languages | - English
- Traditional Chinese
|
Supported Image Sources | Any H.264/H.265 RTSP IP Camera |
Video Sources Configuration |
Region of Interest (ROI) | Setup ROI area for face capture |
Face Capture Parameters | - Minimum capture face size
- Frame capture interval
- Side face filter
- Biggest face only filter
|
Face Verification Parameters | - Minimum face recognition threshold
- Merge recognized faces interval
|
Environmental Adaptability |
Camera Installation Height | - 3.0 meters max. camera height
- Mounting position within 10 degrees inclination
|
Supported Image Modes | Color & Monochrome |
Robust Against | - Large expressions
- Beard and hairstyles
- Eyeglasses
- Caps and scarfs
- Irregular lighting
- Partial face occlusion
|
Supported Devices & Integrations |
I/O Relay | - Moxa E1214
- Advantech ADAM 60XX
- VIVOTEK AO-20W I/O Device
|
IP-to-Wiegand Converter | - Doorcard DCT-4TW26
- VIVOTEK AO-20W Wiegand Box
|
HTTP Notifications | Send customized GET/POST HTTP messages |
Others | Use VAST Face RESTful API |
Persons of Interest Management (POI) |
POI Management | - Enroll person using Web client
- Manage face profiles database through Web Portal
|
POI Authentication | - Using Face as a credential
- Using Card Number
|
POI Groups | - Default groups: VIP, Blacklist, Staff
- User defined person groups
|
POI Batch Enrollment | Supports bulk enrollment using Excel & .JPEG files |
Enrollment Photo | Min. face size = 200 x 200 pixel |
Requirements | 1 MB Max. photo size (.JPG or .PNG) |
POI Profile Information | Name, ID #, card #, job position, email, phone, POI group(s), and expiration date |
POI Reports |
Investigation Reports | Generate face recognition reports and filter events by:- Location
- Time period
- Person’s name
- POI Group
- POI type (enrolled or stranger)
|
Attendance Reports | Generate staff clocking reports and filter events by:- Clock-in time
- Clock-out time
- Stay length
- Late Arrival
- Early Leave
|
Actions Reports | Generate a report for which pre-defined actions were triggered in response to an event |
Export Data | Export results to Excel file |
Trigger Rules and System Actions |
Rule Engine | Rule based engine to trigger single or multiple system actions based on:- Event Location
- Schedule
- Detection time period
- Specific POI profile
- POI face group affiliation
- Person type (enrolled or stranger)
- POI is (not) a member of a group
|
Supported System Actions | - Trigger I/O Relay
- Trigger IP-to-Wiegand converter
- Send HTTP GET/POST command
- WebSocket notification
|
System Users Management |
System Users | Register user accounts using Name, E-Mail address & Phone number |
System User Groups | - System Administrator
- Super User
- Person Manager
|
System Inter-Compatibility |
Supported Products | - VIVOTEK VAST2 VMS/CMS
- VIVOTEK VAST Face Manager/CMS
|
Face Recognition Verticals |
Possible Applications | - Identify potential troublemakers (shoplifters, criminals or any other barred individuals)
- Detect VIP members upon arrival to deliver a tailored experience
- Provide physical entry to commercial buildings, residences, and other rooms
- Contactless staff clock in system
- Verify visitor’s identity
|