A new function was added to DataFromSky: Traffic flow visualization. Instead of collecting vehicle counts from manually defined combinations of gates, one can now get a fully automatically generated overview of the situation. Take a look at the pictures below to compare the old and new ways:
What happens behind the scenes is that the trajectories are grouped according to their characteristics and the resulting trajectory groups visualized according to their average. Needless to say, the DataFromSky team is pleased to bring another of computer science concepts – clustering – into practical use in the field of traffic engineering.
Of course, the trajectory clustering is again only another stepping stone on which one can build further. Even the most rudimentary statistics inside the trajectory groups allow getting at more and more interesting kinds of information. Let’s say… Automatic detection of fastest path, extracting the associated speed, time spent in roundabout etc., comparison with the rest of the group,… As always, there is much space to grow and improve. As we all know – the sky’s the limit!
As the academic year is coming to its last weeks, the DataFromSky protégés which you already heard about will boost the team shortly. So look forward to more news of this kind, and if you have an idea for a function, type of data, that can be extracted from the currently available information, we are ready to listen!Read more
A series of recording happened around Venice in recent days. A big difference from our “traditional” quadcopter-based approach was the vehicle used for carrying the camera: a tethered blimp.
This pilot project involved TAU, TrafficLab and S8 Esseotto.
The blimp visited three crossroads around Venice, recording from height of 40-70 meters a total of 4.5 hours or 270 minutes of traffic footage for analysis. The weather was somewhat disfavourable, with strong gusts of wind. The results turned out excellent, though.
Analysis of the videos is underway.
The number of partners has increased in the recent days by four with TAU, Sky Eye Innovations, University of Palermo and Department of Mechanics, Energy, Management and Transportation Engineering at University of Genova.
TAU is a short for Trasporti e Ambiente Urbano s.r.l. The Milano-based company provides turn-key traffic analysis services, using its helium-filled balloons for recording video and DataFromSky as a subcontractor for processing and analysis.
Sky Eye Innovations
Sky Eye Innovations is a Swedish company based in Stockholm, focusing on production and renting of custom-built UAVs. They will offer video recording services and distribution of DataFromSky.
University of Palermo
University of Genova
Department of Mechanics, Energy, Management and Transportation Engineering at University of Genova becomes another of our academic partners.
DataFromSky has a number of features for vehicle and object classification. Take a look at the detailed list as well as the video – and watch in HD!