JBIG2 Supported by OCR

Warning

This publication doesn't include Faculty of Sports Studies. It includes Faculty of Informatics. Official publication website can be found on muni.cz.

Authors	HATLAPATKA Radim
Year of publication	2012
Type	Article in Proceedings
Conference	CEUR Workshop Proceedings, Volume 921
MU Faculty or unit	Faculty of Informatics
Citation
Web	Full text
Field	Informatics
Keywords	jbig2enc; JBIG2; PDF size optimization; compression; DML; OCR; pdfJbIm; DML-CZ; EuDML
Attached files	jbig2-supported-by-ocr.pdf
Description	Digital Mathematical libraries contain a large volume of PDF documents containing scanned text. In this paper, we describe how this documents can be compressed and thus provide them more effectively to the users. We introduce a JBIG2 standard for compressing bitonal images such as scanned text and we discuss issues if OCR is used for improving the compression ratio of jbig2enc open-source encoder. For this purpose, we have designed API for using OCR in jbig2enc which we describe in this paper together with already achieved results.
Related projects:	Czech Republic membership in the European Research Consortium for Informatics and Mathematics The European Digital Mathematics Library