Online surveillance video archive system

Powerful video data models and querying techniques are required to retrieve video segments from the video archives efficiently. Structure of video data model and types of video queries may change depending on an application. In this paper, we present a video data model and wide range of query types considering needs of surveillance video applications. In the video data model, metadata information, automatically extracted moving objects, events and objects' positions are considered. The query set based on our video data model includes semantic queries, spatial queries, regional queries, size-based queries, trajectory queries, and temporal queries. With the developed web-based program, all of the queries can be processed over Internet and their results can be played with streaming technology. With our developed system, the functionality of surveillance video archives is increased