A Real-Time Traffic Violation Detection System on Highways Using Surveillance Cameras and Message-Oriented Middleware

The rapid growth in the number of vehicles has intensified traffic congestion and increased the frequency of rule violations, which remain a major contributing factor to severe accidents, particularly on highways, where high-speed collisions often result in substantial loss of life and property. Therefore, the early, accurate, and automated detection of such violations is crucial for enhancing traffic safety. This study presents a real-time violation detection system designed to identify critical highway infractions, including reversing vehicles, illegally stopped vehicles in the right lane, and unauthorized presence of pedestrians or animals on the roadway. The proposed framework employs an image processing pipeline integrated with a distributed event-driven architecture. Using a topic-based publish/subscribe messaging model built on Apache Kafka, the system efficiently disseminates detected violation events and corresponding visual evidence to authorized units in real time. The architecture is scalable and extensible, enabling the incorporation of additional violation types and analytical modules with minimal system reconfiguration. Experimental evaluations demonstrated that the system achieved reliable detection accuracy under appropriate camera placement and operational conditions, offering a practical foundation for next-generation intelligent transportation and traffic enforcement systems.