Vitracom AI Solution per Videochannel

Know who arrives, where they go and how long they stay: Vitracom VC-AI turns your existing cameras into accurate counting sensors – in retail, in museums, at trade fairs and events, and anywhere people or traffic need to be counted. Instead of mounting a dedicated counting sensor above every entrance, one compact device processes up to eight RTSP video streams at once – each as an independent virtual device with its own configuration. It detects eight object types: people, faces, strollers, wheelchairs, bicycles, luggage, cars and trucks. A single system therefore covers visitor counting and traffic counting together. Counting lines report directional in and out values per object type; service zones report current occupancy and dwell time. For demographic analysis, VC-AI also estimates age and gender. Use the figures to prove footfall, plan space and staffing, and evidence reach or utilisation rates.

With VC-AI, data protection is built in rather than bolted on: detection, tracking and counting run entirely on the device – with no cloud service behind it. Images are processed in memory only and never stored; detected people and faces are blurred in the console. Only count values and anonymous attributes leave the device – which makes sign-off with data protection officers, works councils and public-sector clients considerably faster.

Integration runs on open interfaces: HTTP push API, MQTT and pre-installed Node-RED for custom workflows involving further IoT or data services. Your counting data feeds straight into your existing counting platform or BI system.
Setup and operation take place entirely in the browser – no client software, in German and English. Start at one site and grow your sensor network stream by stream: a single device replaces up to eight standalone sensors and saves on hardware, installation and maintenance costs.

Additional a compute unit is needed, see additional products.