Traffic monitoring solution
Scyfer has joined as a member of the H2020 SETA consortium. SETA creates technologies and methodologies set to change the way we organize, monitor and plan mobility in large metropolitan areas. The solutions build on large, complex dynamic data from millions of citizens, thousands of connected cars, thousands of city sensors and hundreds of distributed databases.
The SETA challenge
SETA challenged us to understand and model mobility with a precision and granularity that is impossible with today’s technologies. The resulting models will be, thus, used to inform decision-makers how to improve town planning and infrastructure. Last but not least, the results provide support for individual citizens to plan their journeys in a more efficient and sustainable way.
So, in short, we were briefed to build a solution that would better count traffic modalities and classify specific traffic situations.
How did Scyfer help?
Scyfer analyzes CCTV camera image data and fuses information from multiple sources together. In the camera stream, we use machine learning to track cars, buses, motorcycles, bicycles and pedestrians. As a result of this we generate high-quality data on the mobility pattern of the viewed area. This, in turn, provides invaluable data for traffic modeling.
Here are a few more examples of how the traffic monitoring solution works:
Detecting direction of movement
This shows direction of movement based on CCTV camera streams. Further, we can use this as extra information added to induction loops to get more detailed information on traffic flows and turn fractions. This ultimately comes in use for advanced traffic analysis and simulations.
Human detection solution
This video shows an example of human detection in CCTV video, in an indoor environment, using Faster RCNN algorithms. This can also count the number of pedestrians in traffic. Furthermore, traffic control and advanced simulation can also apply this. (The data is coming from the EC-funded CAVIAR project/IST 2001 37540.
Counting cars and pedestrians
An example of counting cars and pedestrians, and the direction of movement, based on CCTV camera stream. This also works as input for advanced traffic modeling. (The data came from the ViSOR repository.)
The European Commission funded SETA as part of the Horizon 2020 (H2020) program under contract 688082 . It started on the 1st of February 2016. It will last for 36 months.
Read more on the SETA project and the 13 participating companies in the H2020 consortium.