Analysis of Traffic Flow Data for Smart City Planning in Port Harcourt

Traffic congestion is a persistent challenge in Port Harcourt's expanding urban area, and data-driven traffic analysis is essential for smart urban mobility planning in South South Nigeria. This study analysed traffic flow data collected from 18 traffic monitoring cameras installed at major intersections in Port Harcourt City and Obio-Akpor Local Government Areas, Rivers State. Camera footage was processed using computer vision algorithms to extract vehicle counts, speed estimates, and vehicle classification metrics at 15-minute intervals over an eight-week monitoring period. Peak hour identification, traffic density heat mapping, and flow bottleneck analysis were performed. Data revealed that morning peak congestion occurred from 7:15 AM to 9:00 AM, with secondary peaks at 12:30 PM and 5:30 PM to 7:15 PM. Rumuola-Rumuobiakani corridor recorded the highest volume and longest queue lengths, with average speeds below 12 kilometres per hour during peak periods. Commercial tricycles constituted 38.7 percent of vehicle counts and were disproportionately associated with bottleneck formation at narrow road sections. Weekend traffic showed 24 percent lower density than weekdays. Computer vision classification accuracy was 91.3 percent for cars but declined to 78.6 percent for motorcycles due to size variation. Recommendations include signal timing optimisation at the five highest-delay intersections, designated tricycle lanes on major corridors, integration of the monitoring system into a Port Harcourt smart traffic management centre, and use of the dataset for urban mobility master plan development by Rivers State Ministry of Transport.

Keywords: traffic flow analysis, computer vision, smart city, Port Harcourt, urban mobility