News

Automated traffic analysis pipeline: a research collaboration between Technical University of Munich and Imperial College London


Professor Constantinos Antoniou and Dr Tao Ma at TUM Chair of Transportation Systems Engineering are working with Drs Panagiotis Angeloudis and Yuxiang Feng from Imperial College London on a unique and timely seed fund project1: This project aims to assess the impact of COVID-19 on the variations of network-level traffic flow patterns in London, with particular emphasis on the effects of the second lockdown. The partners have developed an automated traffic analysis pipeline that extracts key performance indicators for segments of the London road network, such as traffic volumes, vehicle densities, and vehicle, bicycle and pedestrian trajectories.

Findings from this study can provide more insights into the network-level traffic flow patterns, which can help to improve traffic efficiency and safety. "The analysis has leveraged state-of-the-art computer vision techniques, that were applied on public traffic video feeds, that are openly made available by Transport for London," Dr Angeloudis said."The techniques that were developed as part of this project can be easily adapted to other cities with surveillance cameras".

1 TUM and Imperial co-deliver a number of collaborative programs, including the TUM-Imperial Collaboration (Seed) Fund. This program supports collaborations that fall within the theme of data and its application in science, technology, medicine and business.