|
|
|
Images play a key role in the study of space science. Systems that help users find
relevant images will enable learners to more easily use these images for learning tasks,
such as composing a presentation or writing a report.
Project Description
Content-based retrieval is the retrieval of images via
the use of visual attributes. For example, a user can request to see images that
contain the Space Shuttle taking off in the daytime beneath a clear blue sky.
Content-based retrieval is a powerful tool for image retrieval because it allows
users to search for images using image features, such as color, shape, and texture.
Many of the existing content-based retrieval systems have been developed
for expert users such as scientists and computer professionals. It is
difficult for non-specialist users to use these systems because generalist
users lack expert knowledge about computers and information retrieval
systems. Therefore, our research focuses on content-based retrieval
systems much be tailored to reflect the specific needs of non-specialist
users so that these audiences can fully realize the benefits of
content-based retrieval. To accomplish this, system developers must adopt
a user-centered approach to system design. Our focus is also on the study
of user behavior in formulating an image query and evaluating search
results.
The Earth and Space Science Browser (ESSB) prototype
system has over 1,400 space science images. Each of these images is annotated with text
information that explains important details about the image. This information includes
where the images was taken, the subject of the image, and relevant background information.
|
