Improving traffic by way of video analytics
Source: William Pao
Needless to say, municipalities across the globe are turning to intelligent transportation systems (ITS) to make their cities more livable and sustainable. More and more, video surveillance combined with analytics
has become a key component in ITS, helping manage traffic and reduce congestion.
It goes without saying that urbanization has become a rapid phenomenon across the globe, as more and more people migrate from rural areas into cities.
A byproduct of urbanization, inevitably, is increased traffic, which can cause longer commute time, pollution and bad mood among drivers. On tollways, for example, it’s common to see thousands of cars stall in front of manual toll collection booths. When it comes to parking, it’s also common that drivers in large cities spent quite some time circling around to find available spaces.
As a result, municipal operators looking to solve traffic problems increasingly turn to ITS, which leverages various sensors and devices that are integrated into systems to help in this regard. Already in advanced markets, such as Singapore, ITS use cases are abundant, ranging from electronic road pricing, electronic toll collection
, automatic traffic management
systems to smart parking.
In fact, ITS has huge growth potential, with Grand View Research forecasting that the global ITS market would increase grow at a compound annual growth rate of 10.5 percent to 2025, from 2018’s US$25.58 billion.
“High traffic congestion due to rising number of vehicles has contributed to the need for advanced public traffic management systems. Subsequent need for smart vehicles with public-private partnerships is foreseen to be a major growth driver. Numerous initiatives undertaken to deploy advanced traffic management systems such as vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication is also projected to be observed as some of the key growth strategies,” it said.
How video analytics can help
Various components can be used in ITS, ranging from satellites to tag/readers to radars to parking sensors to onboard units in cars. Yet more and more, smart cameras embedded with incident detection analytics, for example, are being used to detect traffic flow or abnormalities and inform operators accordingly.
“Automatic incident detection brings remarkable outcomes, such as picking up pedestrians in crosswalks and alerting traffic controllers to preempt traffic signals to enhance safety for people in intersections; detecting incidents early, enabling traffic operators to implement the proper workflows to resolve problems quicker; and integrating with DSRC broadcast messages and dynamic message signs to immediately warn drivers of real-time safety issues and increase their situational awareness,” said a recent blogpost
Yet the analytics’ function does not stop just at incident detection and report; the data generated by the camera system will also provide valuable insights for municipal administrators to make future planning.
“City planners can use video as a sensor to continuously gather real-time data so they can analyze the flow of traffic. Video as a sensor also helps city planners to develop and implement new policies that make roadways more efficient and safer. Some examples of data collected include road occupancy; counts for vehicles, bicycles, and pedestrians; and average direction and speed,” the post said. “Cameras with machine learning and advanced analytics
learn over time to recognize defined object classifiers, such as separating out vehicles grouped in front of traffic lights. As the cameras learn patterns over time, the accuracy of their vehicle count improves.”
According to the post, IP cameras can also count the number of open spaces in parking lots and transmit the information to parking and video management systems, which can then share information on open spots and alternative locations to park on dynamic message boards, allowing drivers to find available parking quicker.