TY - CONF
T1 - A Video Database Management System for Advancing Video Database Research
AU - Aref, Walid G.
AU - Catlin, Ann Christine
AU - Fan, Jianping
AU - Elmagarmid, Ahmed Khalifa
AU - Hammad, Moustafa A.
AU - Ilyas, Ihab F.
AU - Marzouk, Mirette S.
AU - Zhu, Xingquan
PY - 2002
Y1 - 2002
N2 - The most useful environments for advancing research and development in video databases are those that provide complete video database management, including (1) video preprocessing for content representation and indexing, (2) storage management for video, metadata and indices, (3) image and semantic-based query processing, (4) realtime buffer management, and (5) continuous media streaming. Such environments support the entire process of investigating, implementing, analyzing and evaluating new techniques, thus identifying in a concrete way which techniques are truly practical and robust. In this paper we present a video database research initiative that culminated in the successful development of VDBMS, a video database research platform that supports comprehensive and efficient database management for digital video. We describe key video processing components of the system and illustrate the value of VDBMS as a research platform by describing several research projects carried out within the VDBMS environment. These include MPEG7 document support for video feature import and export, a new query operator for optimal multi-feature image similarity matching, secure access control for streaming video, and the mining of medical video data using hierarchical content organization.
AB - The most useful environments for advancing research and development in video databases are those that provide complete video database management, including (1) video preprocessing for content representation and indexing, (2) storage management for video, metadata and indices, (3) image and semantic-based query processing, (4) realtime buffer management, and (5) continuous media streaming. Such environments support the entire process of investigating, implementing, analyzing and evaluating new techniques, thus identifying in a concrete way which techniques are truly practical and robust. In this paper we present a video database research initiative that culminated in the successful development of VDBMS, a video database research platform that supports comprehensive and efficient database management for digital video. We describe key video processing components of the system and illustrate the value of VDBMS as a research platform by describing several research projects carried out within the VDBMS environment. These include MPEG7 document support for video feature import and export, a new query operator for optimal multi-feature image similarity matching, secure access control for streaming video, and the mining of medical video data using hierarchical content organization.
M3 - Paper
ER -