How to choose a suitable threshold for Binarisation

[易鑫]

Hello,everyone:

   In the tesseract wiki   "https://github.com/tesseract-ocr/tesseract/wiki/ImproveQuality", it says the importance of  binarisation to the final recognition result. I have tired many methods for choose a suitable threshold, but I have not find a very perfect method.

   Does anyone have the experience on choose a suitable threshold for binarisation,please teach me,thanks a lot.

Sorry for my poor english.

[dimuc...@gmail.com]

Hey,

i found this theis that propose a way to calculate the threshold and represent other methods :

https://d-nb.info/989481123/34

i would help.

best regards

[易鑫]

thank you very much

<dimuc...@gmail.com> 于2019年3月7日周四 上午1:26写道：

--
You received this message because you are subscribed to the Google Groups "tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-oc...@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
Visit this group at https://groups.google.com/group/tesseract-ocr.
To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/e7ae47cc-c392-45da-8c6c-909fc712f2be%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Lars Fricke]

Hi,

we use the local Phansalkar binarizer with default values. you can try that very easy in the imagej fiji package. works perfect for all kinds of input, especially with scans with poor lightning and bad contrast. 

there is a nice comparison and yet another method in 

https://www.google.com/url?sa=t&source=web&rct=j&url=https://arxiv.org/pdf/1609.08078&ved=2ahUKEwjasPKLgvLgAhXLJVAKHdrHCV8QFjAAegQIBRAB&usg=AOvVaw1VmuXsg5fEG3hTTkg_7WcZ

regards lars

--
You received this message because you are subscribed to the Google Groups "tesseract-ocr" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-oc...@googlegroups.com.
To post to this group, send email to tesser...@googlegroups.com.
Visit this group at https://groups.google.com/group/tesseract-ocr.

To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/b0db5b19-386a-4d43-9506-7fb85801ef87%40googlegroups.com.

[Lorenzo Bolzani]

I someone wants to try this and is looking for a python implementation here is one:

http://scikit-image.org/docs/dev/auto_examples/segmentation/plot_niblack_sauvola.html

https://github.com/scikit-image/scikit-image/pull/905/files/bb6af8ec723776fc821654847aec04a652f70042

binary_phansalkar = threshold_sauvola(image, w=25, method='phansalkar', r=50)

(I did not try it yet)

Lorenzo

To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/CADXFd_1tz1JhwfD%3DBb9VfP4gpz%3DH_G8nnuDtzpMz9rK7YsAfvw%40mail.gmail.com.

[易鑫]

thank you all.

Lorenzo Bolzani <l.bo...@gmail.com> 于2019年3月8日周五 下午5:06写道：

To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/CAMgOLLxevU0ZyuH7iTBehQUs_YyfnfWwLcUEZeZc0SP_mOJjEA%40mail.gmail.com.