OCR Serbian

[mikolajddd]

Hi,

I want to OCR a document written in Serbian (Cyrillic script), Latin and German. It's a 19th century book. After I start the OCR process the pages are optimized, and then an error appears. It says:

Could not open data file srp_latin.traineddata. Try resetting!

I tried resetting, but the error still appears. To be clear, I don't need Serbian-Latin OCR, but normal, Cyrillic Serbian.

[Stefan Ziegler]

Which language files did you select?

Language files are downloaded from the internet if you do not have them installed locally, so for the download, an internet connection is required.

[mikolajddd]

I have a stable internet connection. I already OCRed documents in Russian, Ukrainian, Belarusian, Polish, and the only error I get is with Serbian.

[Stefan Ziegler]

The file name of srp_latin.traineddata should be srp_latn.traineddata, wothout the i.

[mikolajddd]

Sorry, where is the file supposed to be located?

[Stefan Ziegler]

Please send me your trainDataList.txt file to forum@pdf24.org, which is located in the %LOCALAPPDATA%\PDF24\tesseract\5.4.1 directory.

[mikolajddd]

I dont have the file. PDF24tesseract only contains tessdata folder, LICENSE and tesseract.exe

[Stefan Ziegler]

Do you use the latest version of pdf24 creator? Which version do you use?

[mikolajddd]

11.26.1. Should I reinstall?

[Stefan Ziegler]

That's fine. Try to delete the %LOCALAPPDATA%PDF24tesseract folder and then try OCR again.

[mikolajddd]

After I deleted the folder, there is an error saying something is wrong with internet connection and it can't download the language files. But my connection is fine.

[Stefan Ziegler]

Maybe blocked by a firewall?

[mikolajddd]

I made an exception, still doesn't work

[Stefan Ziegler]

Please send me your trainDataList.txt file to forum@pdf24.org, which must located in the %LOCALAPPDATA%PDF24tesseract5.4.1 directory.

[Author]

Posts