Calamari-OCR / calamari

Line based ATR Engine based on OCRopy
GNU General Public License v3.0
1.05k stars 209 forks source link

missing output data #170

Closed usteiner9 closed 4 years ago

usteiner9 commented 4 years ago

Hi, I ran tests wit the antiqua_modern model and the pred.txt is either empty or contains only one or two letters - anything I miss here?

Tanks Uwe

andbue commented 4 years ago

Hi Uwe, could you please append an image file? Without any further information, I can only guess:

andbue commented 4 years ago

Quoting readme.md: "Currently only OCR on lines is supported." You could cut out lines using ocropus or kraken segmenters or manually create a PAGE file containing line segments using LAREX or Aletheia

chreul commented 4 years ago

exactly, the problem was that you used an entire page as input. please consider the links provided in https://github.com/Calamari-OCR/calamari/issues/169