Research Assistants: Melli Annamalai, Todd Burgette, Joe Ellis, Dana Farr
Sponsors: NASA, NSF, AT&T
The goal of the Global Information Infrastructure is to provide access to distributed stores of information at a reasonable cost. The large size of data items in this global digital library leads to increased processing time during search, retrieval and communication of data. The focus of our research is to address the efficiency issues while performing these operations on image data items distributed over a wide area network.
We are working on efficient techniques for the communication of images. We have developed a tool to quantify the loss in quality of an image during lossy compression. This will determine the quality of image that has to be transmitted so that it satisfies a particular application, and specifically is indistinguishable from the original image to the human eye. This will result in the usage of the network resources only when absolutely required.
We define the concept of equivalence to optimize search and retrieval in globally distributed multimedia information sources. Multimedia data items such as video and images can be represented by an equivalent data item (unlike text) which is an approximation of the original data item and this equivalent data item can be retrieved instead of the original. We are developing techniques to optimize digital library queries based on this concept. Our experimental evaluation uses images from NASA as a case study.