Real-Time Traffic Flow Controls at Signalised Intersections: A Case Study of the Said-Jawda intersection in Karbala

Abstract

As the problem of urban traffic congestion increases, the need for advanced technology and equipment to be introduced for the improvement of traffic control also increases. Traffic problems have worsened last years due to significant increases in the number of vehicles on the roads and the limited availability of resources in the current infrastructure. The easiest way to control traffic lights is to use timer for each phase; however, another way is to use electronic sensors to detect vehicles to supply the required traffic signals for a given cycle. To enable effective traffic flow through a signalised intersection, this study proposes a system for controlling traffic lights that detects vehicles using videos and imaging techniques instead of electronic sensors embedded in the pavement. Four cameras were installed at a height of three metres on the four approaches to a set of traffic lights. The captured images were then analysed using digital image processing based on Python to facilitate vehicle detection and traffic condition analysis. The work is thus divided into two segments: data collection and traffic control solution generation. The results indicated that using the proposed automated system could produce more effective traffic control as compared to SIDRA fixed-time controls, with delay reduced within the range 40 to 90% and cycle length reduced by 30 to 65%.