OCR of historical printings with an application to building diachronic corpora: A case study using the RIDGES herbal corpus

Springmann and Lüdeling describe the results of a case study that applies Neural
Network-based Optical Character Recognition (OCR) to scanned images of books
printed between 1487 and 1870 by training the OCR engine OCRopus on the RIDGES
herbal text corpus.