Evaluating OCR and Non-OCR Text Representations for Learning Document Classifiers.

Markus Junker Rainer Hoch Evaluating OCR and Non-OCR Text Representations for Learning Document Classifiers. 1060-1066 1997 conf/icdar/1997 ICDAR http://doi.ieeecomputersociety.org/10.1109/ICDAR.1997.620671 db/conf/icdar/icdar1997.html#JunkerH97