Manual traffic surveillance can be a daunting task for city administrators as Traffic Management Centres operate a myriad of cameras. Although it is now common to use computer vision for intelligently analysing images of vehicles, challenges remain for understanding the vehicles’ behaviours on the roads.
It’s difficult for the operators to immediately notice, if there has been a car accident on the road that could cause major traffic jams or a group of speeding motorcycles are making a loud bang and causing a nuisance to citizens. These issues can immediately escalate to threaten the safety of citizens and passengers, and become a major source of noise pollution.
