How realistic is static traffic assignment? Analyzing automatic number-plate recognition data and image processing of real-time traffic maps for investigation

• Main Authors: Hamid Mirzahossein, Iman Gholampour, Maryam Sedghi, Lei Zhu
• Format: Article
• Language: English
• Published: Elsevier 2021-03-01
• Series: Transportation Research Interdisciplinary Perspectives
• Subjects: Origin-destination matrix; Data mining; Image processing; ANPR cameras; User equilibrium; Tehran
• Online Access: http://www.sciencedirect.com/science/article/pii/S2590198221000270

Travel demand information in the form of the origin–destination (OD) matrix plays an essential role in studying urban traffic management and network design. The present study takes a novel step toward urban traffic analysis using data mining of processed images of real-time traffic maps as a location-based data model, in which the data were analyzed by software programs such as KNIME and Python workspaces and comparing the results with the conventional traffic assignment results. Thus, we investigated a real-time OD matrix based on the trip-per-vehicle by automatic number-plate recognition (ANPR) cameras for the congestion charge zone (CCZ) of Tehran, Iran. The obtained matrix was assigned to the CCZ transportation network by the convex combination method concerning user equilibrium (UE) condition. The traffic pattern and assignment results were compared to the real traffic data gathered by ANPR, big data analysis and image processing of real-time traffic maps. Considering that the OD based on the trip-vehicle matrix was estimated for vehicle entrances-exits and found to be acceptably accurate compared to real-life conditions, it could be concluded that the UE could not find the practical assignment in 27% of cases in comparison with reality.