search

tesseract-ocr / tesstrain

Train Tesseract LSTM with make
Apache License 2.0
620 stars 181 forks source link

Best training practice #225

Closed dor01 closed 3 years ago

dor01 commented 3 years ago

Hey there,

I try to fine-tune a model with handwriting text lines (png + txt). The makebox give to any character the rect of all the line. And its work but when I use that traineddata i get bad result the model dident find the location of all characters properly.

My question is if I create the box files with currect location of any character its will give a better results? Or if there is other way to get better results

stale[bot] commented 3 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.