Region-Based Detection of Four-Wheeled Vehicle Density on Highways Using Euclidean Distance

Abstract

Traffic congestion in urban areas such as Bandung has become a critical issue that demands intelligent and efficient solutions. This study proposes an image-based vehicle density detection system for four-wheeled vehicles using a region-based approach combined with the Euclidean Distance algorithm. Traffic images are analyzed to calculate inter-vehicle distances based on centroid points, and the results are used to classify traffic conditions into three categories: free-flowing, moderate, and congested. The system is developed using the Python programming language and the Streamlit web interface. Functional testing is conducted using the Black Box Testing method. Experimental results demonstrate that the system can automatically and reliably detect and classify traffic density in real time, offering a practical solution to support urban traffic management.