American Dynamics ADVER12N0N2GDI, ADVER18R5N2GDI, ADVER30R5N2GDI Specsheet

VideoEdge

DEEP INTELLIGENCE NETWORK VIDEO RECORDER

Gain Highly Accurate Security and Business Data with
Pre-Configured Deep Intelligence

The VideoEdge Deep Intelligence Network Video Recorder (NVR) provides the foundation
for powerful deep learning analytics that will open up a portfolio of AI-trained neural network
models for improved security and business intelligence. The hardware offers highly
accurate people counting and tracking from an overhead view by distinguishing between
humans and objects in the camera’s field of view. The analytics engine can also track
movement, loitering, entering and occupancy with confidence.

Ideal for the hospitality, retail and healthcare industries, the NVR is pre-trained with
thousands of images and will work for the majority of environments upon installation.

The victor client and VideoEdge hardware platforms improve the efficiency of security
personnel and daily business operations by creating a powerful video management
solution that allows users to leverage high-performance video streaming, analytics and a
leading-edge feature set for streamlined command and management.

FEATURES THAT MAKE A DIFFERENCE

• A GRAPHIC PROCESSING UNIT (GPU)

OPTIMIZES THE NVR’S ABILITY TO
DELIVER HIGHLY ACCURATE VIDEO
ANALYTICS

• COMBINE WITH THE FULL SUITE OF

ANALYTICS AVAILABLE THROUGH THE
VICTOR VIDEO MANAGEMENT SYSTEM
TO UNLOCK INSIGHTFUL BUSINESS DATA

• FEATURES CAPABILITIES OF THE 2U NVR

SUCH AS VIDEOEDGE TRICKLESTOR,
FRONT-ACCESSIBLE DRIVES AND TRIPLE
STREAMING

www.americandynamics.net

• QUICKLY IDENTIFY SUSPICIOUS

BEHAVIOR AND MOVEMENTS WITH
CUSTOM CONFIGURED ZONES OF
INTEREST

• ANALYZE FOOT TRAFFIC PATTERNS AND

CONSUMER SHOPPING ACTIVITY WITH
PEOPLE COUNTING

• H.265 LIVE STREAMING, RECORDING

AND PLAYBACK SUPPORT MANAGE
HIGHER RESOLUTIONS WHILE REDUCING
NECESSARY STORAGE SPACE

• IDEAL FOR SECURITY APPLICATIONS

AS WELL AS IMPROVING CUSTOMER
SERVICE AND MEASURING THE
SUCCESS OF PROMOTIONS

• COMPLIES WITH THE CYBER

PROTECTION PRODUCT SECURITY
PROGRAM

Hardware Specifications

Specification Description

Camera Type
Max # of Cameras
Deep Intelligence Streams
People Counting
Operating System
OS Drive
Network Interface
RAID Controller
Video Storage
External Storage
Power Supply
Redundant Power Supply
Max BTU
Monitor Interface
Max. Digital Monitors
VideoEdge Client
Video Recording Throughput
Dimensions (W x H x D)

FCC Part 15, Class A; CE: EN55022, Class A; CE: EN6100-3-2; 3-3; ICES-003/NMB-003,Class A; AS/NZS CISPR22,

Regulatory

1

150 Mbps if VideoEdge Client is used

2

Accuracy dependent on environment and may require model tuning

Class A Immunity CE: EN50130-4 CE: EN55024 Safety UL 60950-1 (2nd Ed); IEC/EN 60950-1; CSA C22.2 60950-1,

IP

64

Up to 8

Up to 95% accuracy²

VideoEdge OS built on openSUSE Linux

Single 500 GB

4 x 1 GigE NiCs

Yes; see ordering information

12, 18, 30 TB

iSCSI

400W, 100~240VAC

Yes

1160

1 VGA, 1 DVI

1

Yes

300 Mbps

48.3 x 8.6 x 53 cm (19 x 3.38 x 21 in)

UL 2900-2-3 Level 3

1

Deep Intelligence Capabilities

Color Filtering

Crowd Formation

Direction
Dwell
Enter
Exit
Linger
Object Detection

Queue Length

Tripwire (People Counting)

Hone in on objects or people by searching video based on color attributes, which can be applied to any deep intelligence rule. For
example, quickly find a red truck or a person wearing a green sweater.

Set a region of interest and when a crowd of a certain size forms, an alert can be raised. Improve responsiveness to "flash mob"
and other multiple offender crimes.

Track people moving in different directions for situational awareness and traffic trends.

Draw a region of interest that triggers an alarm when a person within the field is stationary for an extended period of time.

Count objects or people entering a specific area at different times. Generate reports and trigger alarms as needed.

Count objects or people exiting a specific area at different times. Generate reports and trigger alarms as needed.

Determine if people are pacing in certain areas for suspicious or non-permissible periods of time.

Detect people under the camera as they enter the scene and track them as they move around within the field of view.

Sound an alarm when a line gets too long. Analyze data to pinpoint times during the day when lines are empty, short, medium or
long to better manage cashier lines, toll lanes, etc.

Detect when a person crosses a virtual tripwire as well as count people crossing in and out. If the count reaches a given threshold
an alert can be generated.