ABSTRACT
Capturing image data through the use of digital equipment is one thing, aiding users in their attempt to locate and find digital images is quite another. The creation and maintenance of image metadata, or information about images, is crucial to current approaches to image searching, which generally involve entering a string of text that describes the subject or concept required, and querying a database which is expected to hold detailed and accurate textual records regarding the images. In order to provide access to relevant images, good, detailed, systematic and consistent descriptions of their contents and provenance is required. Conceptbased image retrieval is therefore currently dependent on the creation and maintenance of good quality metadata. Additional information can be recorded about other image attributes such as format, creators, provenance or time and place of creation, which is essential information for the technical and managerial upkeep of digital image collections.
