Metadata for Digital Images
Managing a collection of digital image assets is a complex undertaking. Unlike text-based documents, images do not provide headings, captions, or the ability to count word frequency for delivering relevant search results. With examples from stock photo distributors and a case study on an image collection of 3.5 million, this session will address critical considerations for building effective metadata for images. We’ll discuss the particulars of image distribution systems, metadata for rights protection and topic access, and overview primary metadata standards for managing digital images.
David Riecks, project leader for the Stock Artists Alliance Photo Metadata project (http://www.photometadata.org/) will explain the benefits of using embedded photo metadata such as captions, keywords, geographic information, and rights-based information in managing image collections. Emphasis will be on the proper use of existing photo metadata standards that are in wide use, such as IPTC, XMP and Exif, as well as recommended best practices for photographers and for those maintaining image databases.
Jody Apap, Vice President of Operations at A2Z Keywording, will continue with addressing the difficulties of keywording million-image databases, including how to describe complex visual images, how to insure limited and yet accurate and consistent search results, the problems of synonyms and antonyms, and the ambiguities of language. He will also discuss the importance of taxonomies, how to create search-friendly keywords, and give an in-depth look at search functions and criteria.