This project presents a system for real-time vehicle detection and traffic density estimation, utilizing the YOLOv8 model. Leveraging the advanced object detection capabilities of YOLOv8, the system effectively identifies and counts vehicles within predefined regions of traffic scenes, addressing the growing demand for efficient urban traffic management. YOLOv8’s optimized architecture enables accurate multi-scale vehicle detection across diverse environments in real time, providing both high speed and precision. By analyzing traffic flows and calculating vehicle counts per frame, the system can identify peak traffic hours, assess congestion levels, and contribute to enhanced urban traffic planning. Experimental results demonstrate the model’s robustness in handling real-world traffic scenarios, underscoring its potential as a valuable tool for contemporary smart city applications.