Suggestion: Adding an OCR option to capture un accessible web page like Chrome

Re:Suggestion: Adding an OCR option to capture un accessible web page like Chrome (Support)

Thu, 26 Jan 2012 09:19:11 GMT

Thank you for your input, Martin and Thomas! I actually look in this direction too. Experimentally, I connected Tesseract - public domain OCR, now supported by Google - to HyperSnap 7. For some screen fonts it gives very good results, for other - really hopeless. After some more work I'll let you play with it, and I'll also consider training Tesseract myself for screen fonts. For now, see how it recognized the text of Thomas's message:

Re:Suggestion: Adding an OCR option to capture un accessible web page like Chrome h ago 9:] Reply to message
(permalink)
Hello Martin, â€˜
as you may have read in the forum entries, IE and FF modiï¬ed their code some time ago, making it almost 0
impossible to do some decent text snap. Chrome was never actively supported. There are simply too
many browsers and not enough ressources to get into the coding.
OCR would require to license another software, which would add to the price of HS. Additionally there would be more code
to monitor and ï¬x.
MS released an OCR module (leftover from Ofï¬ce 2007) for free recently. It can be combined with scanner drivers. Google it
and maybe you can combine it with HS.
Personally I have no experience with it and made up this tip just now.
Thomas
Lesser Admin
I}
The more things stay, the more they diange the sane.

It's particularly bad in recognizing "fi" or "Fi" letters combination. Some graphics is also mistaken for letters, often foreign letters. I did not try yet to replicate the screen layout text on the ouput, so even if there is a lot of space between some fragments, it's compacted to one space.

Greg

Re:Suggestion: Adding an OCR option to capture un accessible web page like Chrome (TomD101)

Thu, 26 Jan 2012 00:33:02 GMT

Hello Martin,
as you may have read in the forum entries, IE and FF modified their code some time ago, making it almost impossible to do some decent text snap. Chrome was never actively supported. There are simply too many browsers and not enough ressources to get into the coding.
OCR would require to license another software, which would add to the price of HS. Additionally there would be more code to monitor and fix.
MS released an OCR module (leftover from Office 2007) for free recently. It can be combined with scanner drivers. Google it and maybe you can combine it with HS.
Personally I have no experience with it and made up this tip just now.
Thomas
Lesser Admin

Suggestion: Adding an OCR option to capture un accessible web page like Chrome (arjfca)

Wed, 25 Jan 2012 16:56:58 GMT

An idea that I have. Why not adding an OCR to capture text from web page or application where TextSnap is not able to work.

I understand that this is patch, but I would prefer to use only one tool instead of 2 to do all my screen ans text capture. For sure, If tha capture could be enable in all type of application, then there is no need.

Actually, I can't capture Chrome and TWS (Interactive Broker application) windows

Martin

Please provide this additional data:
Exact HyperSnap version in use (from Help/About menu): ____7.11__
Windows version in use: __7 64____