Resource: Using Kraken to Train your own OCR Models
Digital Humanities Now 2019-11-07
Summary:
From the resource: Over the summer of 2019, inspired by the promising results in articles like Romanov et al. 2017, I set out to use the Kraken OCR software on a variety of texts. Kraken, see their website or their repository, is open-source command line software that is capable of reaching accuracy rates in the high nineties for Arabic and Persian printed text. Kraken is not equipped to handle every text – I recommend...