Brown Sloth

@bmoren I would be willing to take this up.

@bmoren I am done with the changes. This might sound trivial but I need to test them, what's the procedure for that?

It could be cause a particular font that is used by your pdf (most probably ZaDb https://fontsgeek.com/fonts/zapf-dingbats-regular responsible for things like radio buttons and checkboxes) is not embedded in the...

Hi. I have tried simulating this issue by trying out the library on: 1. Fake pdf: `echo "Some random text" > broken.pdf` 2. Blank pdf: `touch blank.pdf` 3. Corrupted pdf:...

Working for me: ``` import easyocr reader = easyocr.Reader(['en'], verbose=False, gpu=False) # this needs to run only once to load the model into memory result = reader.readtext('doesnt_see_margin_text.png', detail=0) print(result) ```...

This has nothing to do with easyocr. Probably an issue with pip itself: https://stackoverflow.com/questions/67273590/pip-21-1-cant-import-invalidschemecombination https://stackoverflow.com/questions/67446140/importerror-cannot-import-name-invalidschemecombination-from-pip-internal-e/67508273

@gurudk is right. The following code works: ``` import easyocr reader = easyocr.Reader(['en']) # this needs to run only once to load the model into memory result = reader.readtext('zero.jpg', text_threshold=0.6)...

Seems like a possible issue with the cpu arch/opencv build. Architecture requirements as mentioned here https://github.com/JaidedAI/EasyOCR/issues/646#issuecomment-1050816250 could be an issue if you have an older processor version. You could try...

Working on this:  Anyway if out of the box is not working you could try configuration which can help bring out the decimal. Like: ``` # result = reader.readtext('decimal.png',...

What @howff said, but that might cause a "CUDA out of memory" earlier for your usecase. If so: - If you are processing multiple images in parallel, you can reduce...