Language auto-detection in the Cloud?

  • Last Post 26 May 2014
David posted this 26 May 2014

Dear Supporters

Cloud OCR SDK Documentation ( lists all available languages. I have noticed that Finereader 11/12 lists language auto-detection of the language as feature, the cloud doesn't seem to implement an language 'Auto-detection'. a) Are there any plans making language auto-detection available in the Cloud OCR?

b) Any suggestion how I can work-around this issue? PS: I have noticed when I specify English as language and provide a Japanese Document, the success rate is exactly 0% which should probably encourage anyone not to specify languages that don't apply! Let's assume, I can limit the number of possible language to 10. Would you say there's a chance of success specifying all 10 languages knowing that at least 9 don't apply?


Anastasia Galimova posted this 26 May 2014

Language auto-detection works in FineReader and Cloud OCR SDK in the same way: it is necessary to specify several languages, some of them will be selected automatically. In general the recognition quality is better when less languages are specified.

The recognition quality depends not only on the number of languages, but also on the image quality and languages you use. We recommend to test it on your side.