There are diversified border inspection data including structured data such as alarm records, system logs and operation and maintenance data and non-structured data such as videos and pictures. With continuous increase of the resolutions of front-end monitoring equipment, increasingly longer storage periods of videos and pictures and continuous increase of system scales, the ways to store, share and apply data effectively are more and more important. With “built for practical use” as its goal, big data as its core technology and cloud storage and cloud computation system as its foundation, the system can analyze big data quickly through the cloud computation system. For example, the cloud based video identification technology supports analysis of massive videos and quick extraction of the videos about the customs clearance of designated passengers; the application system realizes further data mining and association and forms effective “files”; the analyzed data and the mined data are finally used for the prediction, prevention and alarming of passengers’ abnormal behavior events, realizing practical application of the border inspection system.