Real-Time Vehicular Traffic Violation Detection in Traffic Monitoring Stream

Traffic violations and offences are becoming more and more serious as the traffic volume increasing, which may bring property damage and threaten personal safety. Existing systems lack of capability to analyze high-throughput traffic monitoring stream and detect various types of violations in real-time. Thus, a real-time vehicular traffic violation detection system is in real demand. In this paper, we design and implement a real-time vehicular traffic violation detection system. Our system proposes a detection algorithm which can discover various types of violations taking place on the roadways as well as in the parking lots. In order to achieve real-time analysis, parallel computing techniques are used in our implementation. An optimization scheme as well as a well-design data structure is proposed to improve the performance of the parallel implementation. Both real data and synthetic data are applied in our experiments. Experimental results demonstrate that our proposed system can discover all the violations from the high-throughput traffic monitoring stream in real-time. Good scalability on multiple threads is also shown in the experimental results.