Support for Application Dictionary in OCR Cloud

  • Last Post 06 June 2014
David posted this 21 May 2014

Dear Support

Does OCR cloud service support use of OCR dictionaries ideally by application ID? Alternatively, can a dictionary word list be submitted with the task?


Order By: Standard | Newest | Votes
Anastasia Galimova posted this 26 May 2014

Unfortunately, in Cloud OCR SDK user dictionaries are not supported. You can only use dictionaries that are included in the predefined languages:

Anastasia Galimova posted this 26 May 2014

Could you please clarify what does it mean "use of OCR dictionaries by application ID"?

David posted this 26 May 2014

Dear Anastasia

I believe, a dictionary (=word list) could increase success rate of business 'slang' words that cannot exist in any other dictionary, such as: - Example: Kolaximundial, Lurakumunda, Pontenarulin - Product codes: MMN-40011, MMN-41113, MMN-41114 (potential issues with capital letter i and the number one) - Perhaps chemistry formulas (not many are around)

In my case, a rather small list 100-300 terms could help to increase the hit rate in recognition.


Anastasia Galimova posted this 06 June 2014

Thank you for the information. Unfortunately, now support of user dictionaries is not in the nearest plans, as usually user dictionary is used for a lot of tasks, and we process tasks separately - so the dictionary should be loaded for each task, it's not optimal.

We have regular expressions in field-level recognition, it could be used as user dictionary, if you recognize text by fields.

Anastasia Galimova posted this 06 June 2014

You can test, if user dictionaries are really helpful for your task, by using the trial version of FineReader.