MultiCam: Video analysis for the automatic recording of traffic flows
Automatic AI-based analysis of multiple video streams for the quantitative recording of multimodal traffic flows in extensive transport systems
1
Background
While we already analyze standard intersections with high precision, extensive areas such as large roundabouts or transit hubs remain a challenge for the industry due to the limited field of view of individual sensors. Capturing these complex infrastructures economically and accurately requires a technological leap that goes beyond processing isolated camera feeds.
2
Objectives
In this project, we are developing an innovative solution for the logical coupling of multiple camera systems. Our primary goal is to implement an analysis pipeline that enables seamless and privacy-compliant tracking of road users across different fields of view. Furthermore, we are creating intelligent planning algorithms that utilize 3D environmental models to automatically determine the optimal number and positioning of sensors required for continuous coverage.
3
Expected Results
The outcome will be a scalable solution capable of fully automating the holistic analysis of traffic relationships at major intersections. For our clients, this eliminates the need for costly manual surveys and provides access to high-precision trajectory data across entire urban areas. Ultimately, this technology provides the essential data foundation for the efficient dimensioning and planning of future traffic infrastructure.

