Mini ITX technical specifications

The Visense CrowdDynamics appliance provides an automatic analysis of passing visitors or crowds. It gives insight about your daily retail business, allows crowd control during events or gives quantified information about the efficiency of your pedestrian and bicycle road infrastructure. The small footprint makes it easy to deploy and scale, and no complicated sensor setup. The Visense CrowdDynamics appliance is an easy way to remotely give real-time information via your IP network.



The software uses a model of the visual appearance of a human. This model is trained with millions of sample images in many different situations. The training, which was performed in our research laboratory, reveals a visual model that is represented by a complex numerical pattern that is used for mathematical comparison of image data. By comparing each position of the image content with the numerical pattern, the system recognizes the presence of a human at each position. This process is repeated 25 times per second on images of 1 Mpixel resolution so that the
walking direction of all people in the image is monitored and people flow (people/hour) is registered. Each tracked individual is counted including information about its position, speed and size and this information is stored in a local database. A web server provides visualization of information graphs and allows third party systems to interface with the appliance to import the metadata information for his needs.

  • Central storage of the metadata from multiple ViSense appliances.
  • Further analysis of the metadata for market-specific information. For example deriving dwelling-time, evacuation time, most busy day of year, etc.
  • Creation of own applications on top of the meta data.
  • Creation of your own market-specific dashboard for services or as a value-added product.
  • Flexible and gives instant information about the current situation suited for a wide variety of applications including crowd control and security.

Fast installation

The sensor is easy to install and similar to installing a surveillance camera. The sensor can be wired with a single
Cat 5e cable using PoE over a distance of up to 100 meters. Simply focus the camera on the area of interest.
No complicated alignment of viewpoint measurement is needed. Configuring of the camera position and setup
of a region of interest are simple steps that are done in the user-friendly GUI.

The information resulting from the passing people is stored safely on a SSD storage device to store up to one year of
counting data. Using a web browser you can access the built-in web server which creases comprehensive visualizations
in bar graphs and line graphs. The web server also allows you to directly access the information using a web service
API (The SDK for this is only available in the Enterprise model).


ViSense CrowdDynamics

Counting width 6 meter 12 meter 18 meter
Maximum throughput 350 pedestrians/minute 700 pedestrians/minute 1000 pedestrians/minute
Maximum object speed 20 km/h 30 km/h 30 km/h
Maximum object density 4 pedestrians/m2
Required light intensity > 3 lux
Image sensor 1280 x 720 pixels
Required frame rate 25 Hz
Video connection  H.264 over RTSP
Operating temperature 0 – 50 ° C
Storage temperature -20 ° – 85 ° C


Counting width 6 meter 12 meter 18 meter
Interface RJ45 – 10/100/1000 BaseT
Cabling Cat 5 (or higher)
Addressing DHCP or Static IP
Time sync NTP
Data delivery protocols IPv4/v6, HTTP, web socket
Supported protocols IPv4/v6, HTTP, TCP, RTSP, UDP, ICMP, DHCP, NTP, DNS
Access level restriction User-based access restriction
Security Password protected web sessions
User interface access Web browser access via build-in web server


Counting width 6 meter 12 meter 18 meter
Operating system Linux
Storage 60 GB SSD (1 year data) 120 GB SSD (1 year data)
Dimensions 8.5 x 8.75 x 1.75 in (21,9 x 22,23 x 4,45 cm)
 Power 19 VDC: max. 60 W, typical 30 W, AC/DC adapter supplied
 Ports USB 2.0, USB 3.0, LAN
 Weight  1,4 kg
Mounting options   Stand-alone, 19” single unit (1U), 19” double unit (1U)
19″ front plate (optional)  1U rack mount
 Enclosure  Black top-grade brushed aluminium front, steel casing

Outline dimensions