Challenge / Goal
The city of Eindhoven wante dto perform temporary traffic measurement to redefine spatial planning and reroute traffic to create a safer situation and simultaniously lower the traffic througput through the center of the city.
Using easy to deploy sensor(s) to get insight in the traffic on a cross ection including pedestrians, bicycles and vehicles acquiring counting, their directions, speed, density, waiting time and queue lenght. Using an overview camera with real-time artificial intelligence in a box enables this.
We adopted state-of-the-art deep learning techniques from publications and implemented a real-time object recognition and tracking system. With special highly parallel hardware we achieved to realize this performance with a small device consuming less than 10W. A webserver allows users to configure the system and offers an API to receive the acquired data. A standard camera interface allows the device to use video streams from existing cameras on the road.
The citizens got a letter stating the purpose of the project and could provide feedback
monitor traffic movement
generate insight on traffic using artificla intelligence
Want to learn more about the lessons learned, financial details and results?Log in