You are viewing limited content. For full access, please sign in.

Question

Question

Using another OCR Engine with SDK

SDK
asked on November 8, 2018 Show version history

Hi LF community,

 

Is it possible to OCR documents with SDK using a custom OCR ? 


 

0 0

Replies

replied on November 8, 2018

Hi Amir,

 

In short, probably yes. But why would you want to do that? Is there a special reason you can't use Omnipage?

1 0
replied on November 10, 2018

Hi Chris, thanks for your reply,

I want to OCR  arabic documents, i checked many externals OCR Engines and i'd like to integrate the most compatible one with LASERICHE SDK in order to OCR Documents directly from LF.

I hope that's clear for you.

0 0
replied on November 12, 2018

Hi Amir,

Makes sense.

Is the language you need not supported? Here is a complete list - http://supportcontent.nuance.com/kbfiles/2674-LangSuppOP12.pdf

1 0
replied on November 14, 2018 Show version history

Hi Chris, thanks for your reply

I guess that Laserfiche uses Omnipage SDK instead of Omnipage Pro, i'm not sure.

here's a comparison table which shows he features of each SDK version (SDK 19.2 and higher, include Arabic OCR Engine). 
https://www.nuance.com/content/dam/nuance/en_us/collateral/imaging/comparison-chart/NDI_344_OmniPage_CSDK_v19-v20.2_Final.pdf

I wonder if Leserfiche will include Omnipage 19.2 or higher in the near future.

1 0
replied on November 14, 2018

Would certainly make a good feature request smiley

1 0
replied on November 8, 2018

The OcrEngine class has a LoadEngine method that allows you to pass in a parameter that specifies which engine to use. However, I don't know anything about how it works, what engines are available or compatible, or what the intended purpose is. There's a GetEngines class that, for me, only ever returns a reference to Omnipage. The client also includes a dropdown that allows you to select the OCR engine to use, but it also only ever has a single entry.

There's a lot more to OCR in Laserfiche than simply retrieving text. The OCR process records pixel locations for each character to facilitate annotations such as highlight and redaction.

1 0
You are not allowed to follow up in this post.

Sign in to reply to this post.