Issue No. 06 - November/December (1998 vol. 10)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/69.738359
<p><b>Abstract</b>—This paper describes the design and implementation of a web-based video database system (WVTDB) that demonstrates our research on video data modeling, semantic content-based video query, and video database system architecture. The video data model of WVTDB is based on multilevel video data abstractions and annotation layering, thus allowing dynamic and incremental video annotation and indexing, multiuser view sharing, and video data reuse. Users can query, retrieve, and browse video data based on their semantic content descriptions and temporal constraints on the video segments. WVTDB employs a modular system architecture that supports distributed video query processing and subquery caching. Several techniques, such as video wrappers and lazy delivery, are also proposed specifically to address the network bandwidth limitations for this kind of web-based system. We also address adaptivity, data access control, and user profile issues.</p>
Video database, video query, video retrieval, video indexing, video browsing, video representation, query cache, World Wide Web, access control.
Haitao Jiang, Ahmed K. Elmagarmid, "WVTDB-A Semantic Content-Based Video Database System on the World Wide Web", IEEE Transactions on Knowledge & Data Engineering, vol. 10, no. , pp. 947-966, November/December 1998, doi:10.1109/69.738359