OCR and pdfs
Dan Beaver
HI,
I have a pdf that I really need to ocr. I am trying to use the Windows ocr via nvda+r. It says it is running the ocr but it comes back with nothing. I know the page of the pdf has text so I am confused about why it isn't OCR ing the thing.
Any ideas?
Thanks.
-- Dan Beaver (KC4DOY)
|
|
JM Casey
Not all text will oCR. How is the text printed?
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Dan Beaver
Sent: February 14, 2019 11:40 AM To: nvda@nvda.groups.io Subject: [nvda] OCR and pdfs
HI,
I have a pdf that I really need to ocr. I am trying to use the Windows ocr via nvda+r. It says it is running the ocr but it comes back with nothing. I know the page of the pdf has text so I am confused about why it isn't OCR ing the thing.
Any ideas?
Thanks.
-- Dan Beaver (KC4DOY)
|
|
Dan Beaver
I do not know since it is a lot of pages that are scanned
images instead of text entered directly into the document.
Dan On 2/14/2019 1:16 PM, JM Casey wrote:
-- Dan Beaver (KC4DOY)
|
|
JM Casey
Yeah, I understand that bit. If it’s poorly printed, or uses a very strange ornamental font, you may not be able to get anything from it through oCR. If it’s images of handwriting you will also get nothing.
Maybe pass the file(s) to someone else and see what they can make of it?
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Dan Beaver
Sent: February 14, 2019 1:33 PM To: nvda@nvda.groups.io Subject: Re: [nvda] OCR and pdfs
I do not know since it is a lot of pages that are scanned images instead of text entered directly into the document.
Dan On 2/14/2019 1:16 PM, JM Casey wrote:
-- Dan Beaver (KC4DOY)
|
|
Brian's Mail list account
If its really bad printing it probably won't work.
toggle quoted messageShow quoted text
Even some faxes wont ocr properly as the text is slightly askew. Brian bglists@... Sent via blueyonder. Please address personal E-mail to:- briang1@..., putting 'Brian Gaff' in the display name field.
----- Original Message -----
From: "Dan Beaver" <dbeaver888@...> To: <nvda@nvda.groups.io> Sent: Thursday, February 14, 2019 4:40 PM Subject: [nvda] OCR and pdfs HI,
|
|
Dan Beaver
Oh, I see what you were asking.
Well, this is a technical manual for a piece of electronics. I doubt it would have too much printing that would be very ornamental or fancy but I could definitely be wrong about that. I'll ask my wife how it looks.
Thanks.
Dan On 2/14/2019 1:56 PM, JM Casey wrote:
-- Dan Beaver (KC4DOY)
|
|
Chris Shook
JM Casey,
Would these restrictions apply to all screen readers and OCR?
|
|
JM Casey
Yes.
toggle quoted messageShow quoted text
OCR is getting better all the time, but it still has lots of limitations.
-----Original Message-----
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Chris Shook Sent: February 14, 2019 2:18 PM To: nvda@nvda.groups.io Subject: Re: [nvda] OCR and pdfs JM Casey, Would these restrictions apply to all screen readers and OCR?
|
|
JM Casey
Hmm, you’re probably right.
Then I don’t know, especially if you were using this function before and it usually works for you.
I’d happily try and do something with it if you want to send it over. I have Adobe pro on this machine which has its own oCR function.
From: nvda@nvda.groups.io <nvda@nvda.groups.io> On Behalf Of Dan Beaver
Sent: February 14, 2019 2:11 PM To: nvda@nvda.groups.io Subject: Re: [nvda] OCR and pdfs
Oh, I see what you were asking.
Well, this is a technical manual for a piece of electronics. I doubt it would have too much printing that would be very ornamental or fancy but I could definitely be wrong about that. I'll ask my wife how it looks.
Thanks.
Dan On 2/14/2019 1:56 PM, JM Casey wrote:
-- Dan Beaver (KC4DOY)
|
|
Lino Morales <linomorales001@...>
Hi Dan. I hate to suggest this on the NVDA list, but have you tried to use convenient OCR with that other screen reader that has a sharky? Good luck.
Sent from Mail for Windows 10
From: nvda@nvda.groups.io <nvda@nvda.groups.io> on behalf of Dan Beaver <dbeaver888@...>
Sent: Thursday, February 14, 2019 11:40:28 AM To: nvda@nvda.groups.io Subject: [nvda] OCR and pdfs HI,
I have a pdf that I really need to ocr. I am trying to use the Windows ocr via nvda+r. It says it is running the ocr but it comes back with nothing. I know the page of the pdf has text so I am confused about why it isn't OCR ing the thing.
Any ideas?
Thanks.
-- Dan Beaver (KC4DOY)
|
|
Gene
If the document is completely blank, is it possible
it is protected in some way and isn't showing anything to the OCR program?
What program are you opening the document in before scanning it?
I don't work with PDF documents much but a document
where there aren't even garbled attempts at recognition makes me suspicious that
nothing is being seen.
Gene
----- Original Message -----
I do not know since it is a lot of pages that are scanned images instead
of text entered directly into the document. Dan On 2/14/2019 1:16 PM, JM Casey wrote:
-- Dan Beaver (KC4DOY)
|
|
Dan,
Given what you have stated the material is, I cannot presume it's private in any way. If you would, please e-mail me a copy so that I can take a look at it. I love the OCR capabilities of a piece of free software, and have offered this information in the past, and will again here:: Free & Good OCR Software for Image Scanned PDFs But, the above being said and offered, there are some scans of ancient documents that are just so bad (all sorts of dots and spots, characters looking like they're melting and changing shape, and more) that no OCR software is going to do a decent job with it. Having the source PDF to look at I can tell you whether this might be the case with it. -- Brian - Windows 10 Home, 64-Bit, Version 1809, Build 17763 A great deal of intelligence can be invested in ignorance when the need for illusion is deep. ~ Saul Bellow, To Jerusalem and Back
|
|
Brian's Mail list account
Yes I tried out that Document channel on Seeing AI on an Iphone. its very choosey even on paper scanning. IE it often says no text recognised but moving the paper or changing the lighting can then make it read. The problem with a picture pdf is that you have a fixed image to start with, so are denied the photographic adjustments altogether.
toggle quoted messageShow quoted text
Brian bglists@... Sent via blueyonder. Please address personal E-mail to:- briang1@..., putting 'Brian Gaff' in the display name field.
----- Original Message -----
From: "JM Casey" <jmcasey@...> To: <nvda@nvda.groups.io> Sent: Thursday, February 14, 2019 8:23 PM Subject: Re: [nvda] OCR and pdfs Yes.
|
|
Dan Beaver
Hi,
Well, I don't have access to that screen reader. I dropped my subscription a few years ago due to the cost. So not an option for me. ;)
Dan On 2/14/2019 5:07 PM, Lino Morales
wrote:
-- Dan Beaver (KC4DOY)
|
|
Gene
You might be able to use a demo.
Gene
----- Original Message -----
Hi, Well, I don't have access to that screen reader. I dropped my subscription a few years ago due to the cost. So not an option for me. ;) Dan On 2/14/2019 5:07 PM, Lino Morales wrote:
-- Dan Beaver (KC4DOY)
|
|