Integrating video cameras for ALINEA on-ramp queue length estimation

This paper studies the performance impact of integrating video camera sensors into existing ramp metering strategies. A significant problem with utilizing vision based sensors in traffic applications is the high regard and strong preference held for inductive loops. However, inductive loops are more expensive to maintain and install compared to video cameras. This paper aims to determine if video cameras can be utilized as an economic alternative to inductive loops, while maintaining existing ramp metering accuracy. To achieve this, several models are developed for video cameras to provide suitable measurements, which are then combined with traffic simulations for ramp metering. These models are then used to evaluate sensor placement combinations for ramp metering algorithms, with results demonstrating minimal performance impact from integrating video cameras compared to existing inductive loop solutions.