The project “Smart Cameras” investigates the possibilities to automize the monitoring process of managed lanes on highways. During peak hours, the shoulder lane is opened for regular traffic. The project is a cooperation between the Dutch Road Authorities, the Traffic Innovation Center and ViNotion. In the current situation, a traffic operator has to manually inspect the shoulder lane to see if it is not obstructed by vehicles or objects such as lost cargo, before it can be opened. In the project, current and newly developed algorithms are used to analyse the video images from existing highway cameras. The algorithms these smart cameras are evaluated to see if they can support or partially take over the routine work of traffic operators.
ViNotion has extended their ViSense detection and tracking technology for the analysis of road traffic. The system analyses the camera images of existing cameras along the highway and fully automatically recognizes all vehicles. In addition, the speed of each vehicle is measured and advanced statistics of the occupation of each traffic lane are generated. In addition, ViNotion has developed a novel algorithm that detects lost cargo (such as bricks, tires or branches). The developed algorithms supports the inspection process by automatically measuring the traffic intensity. This supports the decision making process for opening and closing of the peak-hour lanes.
Early 2017, the technology has been demonstrated to the traffic operators, the management of the Dutch Road Authorities ad the Traffic Innovation Center in Helmond. An extensive validation of the system during the completion of the project has shown that the system is accurate and robust and can effectively support the traffic operators.
For more information, we refer to the project page and the newsitems here and here.
Project: Smart cameras
Nieuws1: Automatic proces more reliable, faster and more predictable
Nieuws 2: Cameras learn to recognize bicycle and jerrycan