Microsoft OCR only english

Help
#1

Hi Robots,

I try to use Microsoft OCR engine on Greek text. In the uipath documentation i read that you can use any language installed on the PC. Although in my pc Greek is the main language the drop down shows only English

1
11920×1080 116 KB

Even when i try to put Greek language in the properties of the activity by hand although it doesn’t give a warning for not recognized language again it doesn’t work for any Greek character

#2

Hi @Poulos_Spyros,

Please write “greek” in ‘Language’ field and try. I haven’t ‘Greek’ language in my laptop so, I can’t verify.

I’m attaching a screenshot, please have a look.

Screenshot%20(35)
Screenshot%20(35)1366×768 87 KB

Thanks & Regards,
Apurba

#3

Thanks but but it doesn’t work. I have already tried this and it gives the same result as if it is set to English.
Does not recognize any Greek character.

1
11466×486 16.6 KB

#4

@Poulos_Spyros Have a look here.Hope it helps.

1 Like
#5

Hi @Poulos_Spyros,

According to the ‘UiPath Documentation’ you should write all letters in small case like “greek”. I found in the screenshot that you wrote “Greek”. However I don’t know whether it will work or not, but please have a try.
Microsoft OCR Link

Thanks & Regards,
Apurba

#6

Thanks Apurba. Again all Greek characters not recognized and depicted as weird symbols.

image
image1920×1080 228 KB

#7

Hey @Poulos_Spyros,

Can you please share a .pdf in ‘Greek’ language?

Thanks & Regards,
Apurba

#8

ΜΙΑ ΚΑΛΗ ΠΑΡΟΥΣΙΑΣΗ.pdf (2.0 MB)

Here you are… Thanks for keep trying

#9

Hey @Poulos_Spyros,

Please try with ‘Get PDF Text’, it will work fine.

image
image634×636 7.3 KB

Thanks & Regards,
Apurba

#10

The majority of documents we have to process are not typed PDFs but scanned documents like the one i attach now. Is there any other way to extract the text out of OCR?2.pdf (43.0 KB)

#11

Hi @Poulos_Spyros,

I have tried few ‘OCR Engines’ along with ‘Microsoft OCR’, but all of these are trying to read .pdf files with ‘English’ language.

Thanks & Regards,
Apurba

#12

@apurba2samanta I think the free version of Microsoft OCR is not supporting to read other languages, try giving a shot using Computer Vision or Google Cloud Vision OCR which has Machine Learning Capabilities, you can get a API key as trail from google or Microsoft azure.