Bilkent University
Department of Computer Engineering


Database Modeling and Integrated Querying for Visual Surveillance in Indoor Environments


Ediz Saykol

Ph.D Student
Computer Engineering
Bilkent University

Automated visual surveillance is a trendy application domain in recent years. Many approaches have been developed on video processing and understanding. In the mean time, some advanced studies have been published on automated visual surveillance technology. Hence, content-based access to surveillance video has become an interesting and important area for researchers. Many approaches have been published that are dealing with automated access to visual surveillance. However, the event models and the content-based retrieval components have significant gaps remaining unfilled. This observation suggests developing a database approach for querying surveillance videos by integrating semantic, spatial and low-level features. This database modeling and integrated querying approach is to be employed for indoor environments, e.g., malls, lobbies, houses, etc. In this talk, the initial steps of this research and some details of the design of the database approach for an integrated querying support are presented.


DATE: November 22, 2004, Monday @ 16:40