| Joyce L. Ogburn Dean and University Librarian University of Utah
| Kenning Arlitsch Associate Dean for IT University of Utah |
| Harold M. Erickson Research Associate University of Utah |
New technologies are making it possible to recover content that has been lost or obscured due to human or natural causes. A team at the University of Utah has developed a process that reveals hidden information in various media, including print, microfilm, and photographs. This presentation is a public pre-launch of retroReveal.org, a project supported and hosted by the J. Willard Marriott Library. The process provides automated, forensic-style enhancement of digital images of varying quality from cameras and scanners, uploaded by anyone. Using different algorithms, dozens of surrogate versions are rendered automatically. Users can then select and annotate the version that best reveals hidden aspects of the image. After processing, users may move images to the public upload gallery, copy to another location, or delete them. In addition, retroReveal.org is a community-oriented site that facilitates scholarly collaboration on interpretation of results.
The beauty of this process is the ability to upload and process any digital format. It is an accessible, inexpensive, and highly effective approach to the problem of revealing hidden information. Though originally targeted at archivists, curators, and conservators, during the alpha phase, the retroReveal algorithms have proven useful in archaeological and other scholarly applications. Examples include recovering content from greatly overexposed microfilms of objects that can no longer be accessed; reading through endpapers to a vellum letter used as a book’s sewing support; recovering a composer’s water-damaged instructions to a publisher concerning musical details of a score; enhancing a poor-quality, low-light aerial photograph that revealed a major archaeo-astronomical complex; visualizing weathered pictographs/petroglyphs; and reading exposure-faded Oregon Trail axle-grease messages on stone. The process also has shown the potential of improving optical character recognition (OCR) to create machine-readable text from image files.
This presentation will include a demonstration of the process as well as discussion of future directions in developing a community of users and contributors.