A Comprehensive Image Processing Suite for Book Re-mastering
ICDAR '05 Proceedings of the Eighth International Conference on Document Analysis and Recognition
| Hi-index | 0.00 |
More than 32M unique book titles are available in US libraries, but Amazon, the biggest retailer, had only 1.2M unique titles available for sale in 2004. Currently there is an effort underway by public libraries, universities, the Open Content Alliance, Google and others, to non-destructively scan these 32M unique books and make them available for on-line viewing and search. Twenty percent (6.4M) of the 32M titles are out of copyright and out of print. A publisher estimates that an average of 40 copies of each title can be sold per year if they could be made available for sale. This long tail opportunity represents a several billion dollar market with the right cost structure. To address this long-tail book market we need to take the cost out of several parts of the value chain: automatic book preparation to minimize publishing setup costs, print-on-demand to remove warehouse and waste costs, and web 2.0 techniques to minimize marketing costs. We have created this system with several partners based on HP technology, and available as an incubation business.