General park ecological algorithm machine

General Park ecological Algorithm integrated machine is a soft and hard integrated ecological product jointly launched by Computer Energy and ecological partners, focusing on the park field, using BM1684 powerful AI chip, built-in park/campus/community and other scenarios required a variety of special AI algorithms, effectively ensure the safety of the park/campus/community and other personnel, equipment assets, one-click deployment, effective in time. To provide customers with the most competitive intelligent solutions.

SE5 16-AA1-21
SE5 8-AA2-21
Scene Specific Algorithm

Rich algorithms, involving video full target analysis, behavior analysis, perimeter/fireworks warning and other algorithms, on-demand call, free combination, AI to enable general park intelligent management

Flexible Deployment

Free placement in all kinds of environments, Completed in 5 minutes. Support RTSP standard protocol. Support through ONVIF protocol scanning equipment. Support national standard stream access

Rich In Function

Global information overview, data statistics, video preview, intelligent search to map, early warning layout control, HDMI direct, event video optional intelligent analysis management platform.

Computational Power Surging

8/16 channel 1080p HD video intelligent analysis , Each video supports video full target analysis and 2 other algorithms at the same time

Deployment Mode

deploy-img deploy-img

Application Scenarios

Smart Park

Manufacturing

Chemicals Park

Commercial Complex

Smart Community

Smart Campus

Smart Park

Manufacturing

Chemicals Park

Commercial Complex

Smart Community

Smart Campus

Typical Case

Shanghai Harbor Community Project
Shanghai Lingang community, as an important urban area, has a large number of people and vehicles, and community management has become an important task. In order to improve the safety management level of the community, the general park integrated machine scheme is introduced. At key locations in Lingang communities, such as community entrances, parking lots, public squares, etc., multiple computing machines have been deployed. These devices are equipped with a variety of algorithms, including people, vehicles, unstructured object recognition, people gathering monitoring, fence climbing detection, area intrusion detection, personnel retention detection and face recognition. The all-in-one algorithm monitors the scene in the community in real time through the camera. Once an abnormal event is found, such as personnel entering the restricted area, personnel gathering, climbing over the fence, etc., the system will automatically trigger an alarm, upload the collected data and alarms to the public security image library, and provide real-time community safety information for the public security department. The inclusion of this successful case in Shanghai's five-year action Plan for AI-enabled communities has further promoted the application of intelligent technology in community management to provide a safer and more convenient living environment for urban residents.
example img

Product Specification

SE5 16-AA1-21
SE5 8-AA2-21

Processor

BM1684

Main control CPU:8-core ARM A53 @2.3GHz

Peak computing power:10.6 TOPS(INT8)/1.3TFLOPS(FP32)

Operating system

Linux system

Video decoding type

H.264 & H.265

Data acquisition equipment

The device must support the RTSP standard protocol, ONVIF protocol, or GB28181 protocol

Intelligent parameter

Video parsing

Support 16 1080p HD video intelligent analysis
Each video can support 14 algorithms at the same time

General campus algorithm

Total target analysis

Motor vehicle attributes, license plate recognition, non-motor vehicle attributes, human body attributes, face attributes, tricycle attributes, target characteristics

Behavior analysis

Wandering, leaving, gathering, talking on the phone, smoking

Alert algorithm

Trip wire intrusion, area intrusion, channel blockage, vehicle congestion, area statistics, illegal fireworks detection, electric vehicle detection, garbage detection

Software function

Abnormal alarm

When an exception occurs, both the alarm and message are recorded

Event management

Abnormal events can be uploaded or saved to a specified location, flexible retention time, and provide access to the function

Real-time preview

Preview surveillance videos online in real time. A maximum of 9 videos can be previewed simultaneously

Memory

Standard configuration

12GB LPDDR4x 128bit 64GB/s

eMMC

Standard configuration

32GB

Interface

Network interface

10/100/1000Mbps self-adaptation*2

External interface

Support USB/HDMI/SATA/RS-232 / RS-485 / TF, LTE optional

Power interface

Dc 12V

Operating temperature

Temperature range

-20°C ~ 60°C

Power dissipation

Typical value

~20W

Structural dimension

Length * width * height

188 * 148 * 44.5 (mm)

Class of protection

Active heat dissipation IP30

Processor

BM1684

Main control CPU:8-core ARM A53 @2.3GHz

Peak computing power:10.6 TOPS(INT8)/1.3TFLOPS(FP32)

Operating system

Linux system

Video decoding type

H.264 & H.265

Data acquisition equipment

The device must support the RTSP standard protocol, ONVIF protocol, or GB28181 protocol

Intelligent parameter

Video parsing

Support 8 1080p HD video intelligent analysis
Each video can support 3 algorithms at the same time

General campus algorithm

Total target analysis

Motor vehicle attributes, license plate recognition, non-motor vehicle attributes, human body attributes, face attributes, tricycle attributes, target characteristics

Behavior analysis

Wandering, leaving, gathering, talking on the phone, smoking

Alert algorithm

Trip wire intrusion, area intrusion, channel blockage, vehicle congestion, area statistics, illegal fireworks detection, electric vehicle detection, garbage detection

Software function

Abnormal alarm

When an exception occurs, both the alarm and message are recorded

Event management

Abnormal events can be uploaded or saved to a specified location, flexible retention time, and provide access to the function

Real-time preview

Preview surveillance videos online in real time. A maximum of 9 videos can be previewed simultaneously

Memory

Standard configuration

6GB

eMMC

Standard configuration

32GB

Interface

Network interface

10/100/1000Mbps self-adaptation*2

External interface

USB3.0 *2 / HDMI *1 / TF *1 / Audio *1

Power interface

Ac 220V

Operating temperature

Temperature range

-20°C ~ 70°C

Power dissipation

Typical value

-18W

Structural dimension

Length * width * height

185 * 170 * 43.6(mm)

Class of protection

Passive cooling IP40