Self-Organizing Maps
Metadata Extraction from PDF Papers for Digital Library Ingest
ICDAR '09 Proceedings of the 2009 10th International Conference on Document Analysis and Recognition
Table of contents recognition for converting PDF documents in e-book formats
Proceedings of the 10th ACM symposium on Document engineering
Conversion of PDF Books in ePub Format
ICDAR '11 Proceedings of the 2011 International Conference on Document Analysis and Recognition
| Hi-index | 0.00 |
The faithful visualization of historical documents on e-book devices and tablet computers is addressed in this paper. To this purpose, digitized books should be converted to re-flowable formats where the characters are easily re-sized. This is accomplished by first analyzing the document to extract the characters that are then clustered and replaced by prototypes. The prototypes are represented as SVG objects and then arranged in the proper position in the converted document. Among other applications, the proposed conversion can be used to allow visitors of archives and exhibitions to easily browse and consult historical documents on dedicated devices or on personal mobile devices that support standard re-flowable formats. The system is quantitatively tested on the well known UW-I dataset by computing OCR errors on the original images and on the reconstructed ones. The visual rendering of historical documents is evaluated on a digitized book of the XIX-th Century.