Digitization and Virtual Restoration of Deteriorating Photographic Materials Eastern Kentucky University Landon, George V.
The goal of this project is to develop a novel method to digitize and virtually restore rapidly deteriorating cellulose acetate film negatives. Working in close collaboration with the University of Kentucky?s Center for Visualization, the Principle Investigator at this primarily undergraduate institution will conduct research in to develop advanced computational models for deteriorating images on unstable and disintegrating media. A successful solution will produce a framework to non-destructively acquire and preserve vast collections of photographic records created from the early 1900s to the 1950s. At the same time, it will provide economic alternatives to memory institutions and research institutions that rely upon the content of the photographs for a variety of scholarly purposes.
During the first half of the 20th century, most photographs were taken on safety acetate film. The composition of the film negatives guarantees that they will deteriorate beyond usable condition even when preserved. If the negatives are caught early enough, it is possible to physically restore them; however, this is time and cost prohibitive and for many negatives not even possible due to their advanced stages of deterioration. This project has led to the development of a complete pipeline for scanning the deteriorated negatives and producing a digital copy that has been virtually restored. This includes the estimated original photographic content, before deterioration made this impossible with normal scanning techniques, and removing distortions due to the change in shape of the negatives. Results of this non-destructive restoration pipeline have been presented at the Library of Congress and professional historical document venues. As part of the findings of this project, a new 3D scanning technique has been developed for materials that can transmit light. Eastern Kentucky University has begun the process to protect this intellectual property and commercialize it. Additionally, methods have been developed to identify and group materials based on their light interaction properties. This has led to new applications in areas like document forensics. During the award period 13 undergraduate computer science students have been directly involved in the project. They have developed algorithms and software while testing various aspects of the restoration system. 8 of these students come from economically distressed counties in Appalachia.
