SBIR Phase I: Book Scanner Technology for Universal Access to Digital Books

This Small Business Innovation Research (SBIR) Phase I research project seeks to develop a low-cost high-speed automatic book scanner to digitize vast libraries of bound printed matter. Despite the ubiquitous use of computers to create modern publications, the majority of literature in circulation remains exclusively outside the digital domain. This renders large volumes of books to be inaccessible to search engines and digital libraries. The vast majority of books that are now converted from the printed page to digital format are scanned manually. This is a slow, tedious, and costly process that is hinders the ability to scan the millions of books by several companies that provide books and content in digital form. For people with physical disabilities, the digital version of books can enable them to reach beyond the mechanical limitations of page turning and the visual limitations of small fixed-sized fonts. Digitizing traditional books is also of practical importance for legal fair-use purposes by students, researchers, and all people who yearn for a digital library that can benefit from portability and search engines. It is anticipated that legal personal use of book scanning by corporations, libraries, government agencies, and high-end consumer markets will have broad impact on education, research, and commerce. The objective of this project is to develop a robust solution at less than one tenth the price of competing products. The proposed device addresses two core problems in book scanning -- automatic page turning and page flattening. A novel turnstile design is introduced to hold the book open flat, as well as to move a page from one side of the book to the other. The analysis of bending stiffness, friction, and contact pressure will enable the input to the image acquisition system to fuse input images having minimal distortion. Feasibility will be determined by measuring the registration (alignment) error of the imaging system and determining if sub-pixel accuracy is achieved in registering multiple page images together during the image fusion operation.