Automated Traffic Video Analysis with a Modular Computer Vision Pipeline and SQL-Based Analytics

The increasing volume of traffic video data demands efficient methods for extracting actionable insights to improve road safety and traffic management. This research presents an automated computer vision pipeline designed to analyze traffic video footage by integrating state-of-the-art object detection models (YOLOv8, YOLOv9, YOLOWorld) with BoT-SORT tracking. Detected entities such as vehicles, pedestrians, and traffic signals are stored in a structured MySQL database along with positional and GPS data. The framework enables quantitative event detection, including jaywalking, red-light violations, and close-proximity incidents, through SQL-based querying. Unlike generative AI models, which are computationally intensive and offer qualitative results, this approach delivers efficient, accurate, and structured analysis accessible to data analysts. Experimental results show high detection accuracy, with precision and recall exceeding 85%. This work provides a scalable solution for traffic video analytics, facilitating applications in urban planning, traffic enforcement, and transportation safety.