/historic-newspaper/README

 1This is a pretty hard document, a scanned, warped historical newspaper
 2page.  It's mostly here as a test case to see how we can improve OCRopus 
 3in the future.
 4
 5The script illustrates how to adjust the layout analysis parameters
 6in ocropus-gpageseg for these kinds of documents. Note that there are
 7some layout analysis errors.
 8
 9Better character recognition performance will require retraining models
10on historical books and newspaper prints (the current models are trained
11on modern scanned documents only).