AN ANNOTATION ENGINE FOR SUPPORTING VIDEO DATABASE POPULATION

Segmentation, video data modeling, and annotation are indispensable op erations necessary for creating and populating a video database. To su pport such video databases, annotation data can be collected as metada ta for the database and subsequently used for indexing and query evalu ation. In this paper we describe the design and development of a video annotation engine, called Vane, intended to solve this problem as a d omain-independent video annotation application. Using the Vane tool, t he annotation of raw video data is achieved through metadata collectio n. This process, which is performed semi-automatically, produces tailo red SGML documents whose purpose is to describe information about the video content. These documents constitute the metadatabase component o f the video database. The video data model which has been developed fo r the metadata, is as open as possible for multiple domain-specific ap plications. The tool is currently in use to annotate a video archive c omprised of educational and news video content.