qurator-spk / sbb_binarization

Document Image Binarization
Apache License 2.0
67 stars 14 forks source link

Why is --patches not the default? #45

Closed mikegerber closed 1 year ago

mikegerber commented 1 year ago

The README says:

Note In virtually all cases, applying the --patches flag will improve the quality of results.

Why is it not the default? Why no --no-patches option instead?

mikegerber commented 1 year ago

FWIW, in the OCR-D CLI, use_patches=True is the default and there does not even seem to be an option to turn it off: https://github.com/qurator-spk/sbb_binarization/blob/master/sbb_binarize/ocrd_cli.py#L109

cneud commented 1 year ago

Full story: https://github.com/qurator-spk/sbb_binarization/pull/5/commits/eece89b9acd40808d98da98093a942518ff4dc5c

TL;DR > ...this parameter should be set "True", otherwise model will see image at once and the quality will be terrible

My vote goes towards always using patches and completely getting rid of the CLI option.