using a custom character set?

  • 961 Views
  • Last Post 20 October 2014
mac21 posted this 16 October 2014

Hello,

is it possible to specify a custom character set by myself? im working with the cloud OCR SDK in Eclipse (java)

i want to use A-Z a-z 0-9 ,.!?

there are some incorrect characters recognized, like ' or ยด or `

Thanks in advance,

Mac

Order By: Standard | Newest | Votes
MFu posted this 17 October 2014

Hallo Mac,

If you use the processFields Method in the Cloud OCR SDK, then you can specify the character set that is used for recognition.

http://ocrsdk.com/documentation/apireference/processFields/ http://ocrsdk.com/documentation/specifications/regular-expressions/

Currently the character restriction in Cloud is only implemented for this scenario.

In the ABBYY SDK family FineReader Engine you have much more granular options to influence the the recognition settings.

Best Regards, Michael

Some related Articles http://www.abbyy-developers.com/en:tech:tasks:zonal_field-level-ocr http://blog.ocrsdk.com/field-level-ocr-what-is-it-for/

  • Liked by
  • mac21
mac21 posted this 20 October 2014

Hi Michael,

thanks for your response. Oh, I did not knew that. For my project I only use "processTextfield", so there is currently no way to specify the charset. Okay never mind, I will remove invalid characters in Java. Just a simple "ReplaceAll('#', '');" for example.

Thanks for your help :)

Close