Hi Ronan and all,
I scanned most of Ó Dónaill (in the eSpeak folder) using Omnipage, set
to Italian. I bought the software in 2003/4, and perhaps it wouldn't
do the same now.
I had to dispose of the scanner because of a driver issue when things
went to Windows 7, and don't know what I've done with the software,
since.
In VVI we're also looking at another angle. Since Irish language
material has to be subsidized by the State, generally speaking, it
needs to be accessible equally to all. Ultimately, all Irish material
produced with State subsidy should be accessible with no third party
software or workarounds, and should be tagged as irish etc. Well, of
course, we'd still need NVDa, but we shouldn't need anything else.
Robbie
On 10/24/20, Ronan McGuirk <ronan.p.mcguirk@xxxxxxxxx> wrote:
Hi all,==========================================================The vicsireland
Just a quick update on FineReader online.
You can register and get an initial 10 pages scanned and a following 5
pages per month for free. After that you need to subscribe for extra
capacity.
I attempted to OCR an 8 page pdf Irish language document.
Initially the interfaceemed very accessible, but the selection of the
document language was a bit confusing. I believed I had selected Irish
but it turned out that I had selected English. The selection process
is unclear. Of course, when it scanned 8 pages as English, the result
was very poor, virtually all the sínte fada (acute accent signs) were
ignored.
I then had 2 pages left. Eventually I managed to identify the document
language as Irish and succeeded in carrying out the scan.
My initial impression is that the Irish Language OCR quality of
FineReader online is good, better than the result I got with
pdfconvertonline.com. The result from pdfconvertonline.com mixed up
the order of a few words. This may be a pdf layout issue rather than
an OCR issue, but FineReader was using the same pdf input file.
Regards,
Ronan
On 21/10/2020, Ronan Mcguirk <ronan.p.mcguirk@xxxxxxxxx> wrote:
Hi Flor,The vicsireland mailing list
Thanks for that. Yes, I have heard good reports on fine reader.
Interestingly, there is a free online service available. It only does a
limited amount of scanning but I might give it a try to see what the
quality
is like.
Thanks,
Ronan
Sent from my iPhone
On 20 Oct 2020, at 20:07, Flor Lynch <florlync@xxxxxx> wrote:
Hi Ronan,
FineReader supports Irish-language OCR. (OmniPage calls their equivalent
OCR support 'Gaelic', but I'm not sure it's as good.) I use Kurzweil
1000
- which has earlier versions of both of these as OCR engines - to scan
and
recognise bilingual documents (election/referendum literature, and the
like), with OCR in a settings file set to both English and Irish. The
results are accurate and readable - at least when using refreshable
Braille. As K1000 apparently hasn't been updated for a few years now,
and
for a cheaper price, you might consider Abbyy FineReader as a
standalone,
about the accessibility of which several individual blind/vi users have
expressed their satisfaction.
-----Original Message----- From: Ronan McGuirk
Sent: Tuesday, October 20, 2020 7:25 PM
To: fb-exchange ; Ilka Stäglin ; Lina Kouzi ; vicsireland
Subject: [vicsireland] Irish Language Scanning options
Hi all,
Just wondering if anyone has any recommendations for the best approach
for Irish language OCR.
You will be aware that using an English language scanning solution
produces very poor results for Irish text. I have been on the
receiving end of files that were effectively unusable.
As far as I am aware, the Microsoft Windows builtin OCR solution, that
integrates with NVDA, does not support Irish. Or, at any rate, I
haven’t got it to work yet, and it seems that the Irish Language files
from Microsoft that it uses don’t support scanning.
Recently I have scanned documents with a flatbed scanner to a pdf
image file. I have then converted this to a searchable pdf file with
www.pdfconvertonline.com. This online service has an Irish Language
OCR option and the results are encouraging.
Just wondering if there are any better options out there.
Any suggestions much appreciated,
I am considering Tesseract, though it seems that it doesn’t support
pdf files directly.
Best regards,
Ronan
====================
The vicsireland mailing list
To unsubscribe at any time send a mail to:
vicsireland-request@xxxxxxxxxxxxx
with the word "unsubscribe", without the quotes in the subject of the
message.
To contact the moderator send mail to:
tim.j.culhane@xxxxxxxxx
For mor information on the Visually Impaired Computer Society visit:
http://www.vicsireland.org
===========================================================
The vicsireland mailing list
To unsubscribe at any time send a mail to:
vicsireland-request@xxxxxxxxxxxxx
with the word "unsubscribe", without the quotes in the subject of the
message.
To contact the moderator send mail to:
tim.j.culhane@xxxxxxxxx
For mor information on the Visually Impaired Computer Society visit:
http://www.vicsireland.org
To unsubscribe at any time send a mail to:
vicsireland-request@xxxxxxxxxxxxx
with the word "unsubscribe", without the quotes in the subject of the
message.
To contact the moderator send mail to:
tim.j.culhane@xxxxxxxxx
For mor information on the Visually Impaired Computer Society visit:
http://www.vicsireland.org