Re: Problem with windows10 OCR and NVDA


Kenny
 

Would be cool if Microsoft would add the ability to send an entire Image/file/PDF to the Windows OCR module in a single step. Believe JAWS already has a feature like that with its OCR facility. This way the entire file could be easily read after the OCR has been processed.


On 2/25/2019 9:01 AM, molly the blind tech lover wrote:

Hi. In my experience it only recognizes a single screen. I needed to use it to read a chapter of my textbook and it only recognized a few sentences. It also only recognized one slide at a time when I needed to use it to access a powerpoint.

 

From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of marcio via Groups.Io
Sent: Monday, February 25, 2019 8:44 AM
To: nvda@nvda.groups.io
Subject: Re: [nvda] Problem with windows10 OCR and NVDA

 

Good question, I would ask the same.

Em 25/02/2019 10:44, Gene escreveu:

does that mean that the NVDA OCR feature only can recognize a single screen or can it recognize an entire document?  I haven't played with this to any extent.  In Notepad. if you open a document, it is all one single object and can all be read with the screen review keys so you aren't limited to one screen when in object review.  What about a document you use the OCR feature on? 

 

Gene

----- Original Message -----

Sent: Monday, February 25, 2019 7:13 AM

Subject: Re: [nvda] Problem with windows10 OCR and NVDA

 

It seems the keys to perform such a command using the laptop layout are NVDA+Shift+up arrow.
I found it by going to the keystrokes reference.

By the way, Rui, thanks for the clarification :)
I'll play with it taking this advice into account and see what I can get.

Em 25/02/2019 09:50, Nevzat Adil escreveu:

"If you want to perform the OCR to  all the window, you will need to
move the navegation object to the window itself, pressing
NVDA+NumPad8..."
 
How does this apply if using a laptop instead of desktop?
 
On 2/25/19, Rui Fontes <rui.fontes@...> wrote:
Hello!
 
Please, remember one thing...
 
When you perform the OCR, using NVDA in Windows 10, pressing NVDA+R, the
OCR is applied only to the object focused by NVDA...
 
By instance, if you are in a dialog type, "save as" and focused in the
Save button, the OCR result will be only "Save", the text of the focused
object, the Save button.
If you want to perform the OCR to  all the window, you will need to move
the navegation object to the window itself, pressing NVDA+NumPad8...
 
Regards,
 
Rui Fontes
 
 
Às 11:54 de 25/02/2019, marcio via Groups.Io escreveu:
Just adding
While with this message opened I tried to OCR just for fun, seeing what
I would get.
What a surprise, the OCR worked just like a charm.
Weird!
 
Cheers,
Marcio
Follow or add me on Facebook <https://facebook.com/firirinfonfon>
 
Em 25/02/2019 08:48, marcio via Groups.Io escreveu:
Very same here!
And I'm also using the latest NVDA Alpha with Win 10 1803.
Well I'm not using the latest at all because I need to update to the
very latest so to speak which showed up just now. I'll do it sometime
later, I guess.
However, the message I get is absolutely the same and no, I don't
think just updating to the latest Alpha will fix it since aparently
the folks on the NVDA dev side weren't aware of it.
Let's wait...
 
Cheers,
Marcio
Follow or add me on Facebook <https://facebook.com/firirinfonfon>
 
Em 25/02/2019 08:31, Mallard escreveu:
Yes, here too. Something must be broken, because I've opened a pdf
that is perfectly visible with any pdf reading apps, including Edge.
 
I tried opening with Sumatra pdf, which only shows the image and not
the text. Usually, when I do that, I can view the page with Inw10
OCR. Now it keeps saying
 
 
"Content not visible" (contenuto non visibile in Italian).
 
 
No idea what could have happened.
 
I'm using the latest Alpha of NVDA and Windows 10 1803.
 
 
Ciao,
 
Ollie
 
 
 
 
Il 25/02/2019 11:57, marcio via Groups.Io ha scritto:
Same here. I don't have Jaws to test with, though, so don't know if
it would be different.
Anyone to help us? :)
 
Cheers,
Marcio
Follow or add me on Facebook <https://facebook.com/firirinfonfon>
 
Em 25/02/2019 06:15, Aravind R escreveu:
Dear friends,
 
I am trying to do OCR with pdf using NVDA+r in windows10.
Wherever i try, it is saying "Recognising result document" but
nothing
is displayed and its working well with jaws's built-in ocr.
Kindly guide where changes to be made to make work with windows10 OCR
 

 

 

Join nvda@nvda.groups.io to automatically receive all group messages.