Design and Realization of Intelligent Public-Transport Flow Monitoring System Based on Big Data

An intelligent public-transport flow monitoring system based on big data is designed and realized so that the increasingly serious bus overloading problem during holidays and festivals can be resolved. The system can achieve intelligent identification and remote inspection of persons loaded in an inter-city bus to prevent overloading. First, system design and architecture are introduced. Second, ways of video key frame extraction, face detection combining Spark with Opencv, and small file merger with MapFile, etc. are resolved under Hadoop and Spark environments. Finally, the system is proved to have good efficiency and expandability through experimental comparison.