One important element of the Digital Library effort is the digitization and archiving of audio visual media for online access by a wide variety of users. In the last few years there have been dramatic improvements in the cost and availability of peripherals that allow the digitizing of this material. The real challenge is to be able to store and catalog this information in a reliable way that supports effective queries and delivery of the right material to the user.
To address this challenge, the Center for Information Systems Optimization (CISO) and the Department of Electrical Engineering have developed a visual media archiving tool called Content. This software runs on standard Windows and NT computers, is scalable and extendible, able to handle millions of media objects efficiently, can be configured to support custom metadata, and is 100% Web-compatible.
The Content software package* is being used to make online multiple graphical collections available to the University community. This, in conjunction with an Intel equipment grant, will provide the initial infrastructure for the Digital Library. One of the most exciting aspects of Content is its ability to allow archivists to define rich and complex metadata to enhance the searching of very large image collections. In addition, these searches can be on indexes across multiple servers, linking diverse archives into a virtual, single collection for the user. Finally, the performance of Content allows the user to retrieve one of a few million images in less than a second.
We have formed a very productive collaboration among CISO, the University Libraries—including the Health Sciences Libraries and Information Center—and CARTAH to begin development of an integrated and well-indexed set of visual media databases. Future versions of Content will incorporate speech recognition, video indexing, and image processing functions. Our goal is to develop Content as a user-friendly, multimedia archival tool.
*Content requires a Java-enabled browser such as Netscape version 3.0 or higher.