OCRdesktop 3.0 Released
Linux for blind general discussion
blinux-list at redhat.com
Sun Jan 16 13:44:50 UTC 2022
Howdy Rastislav, Howdy List,
thank you very much for the nice feedback. Hearing stuff like that is
i was very busy the last couple of days :). Not sure if you know, there
was a small tool named OCRPDF born out of the OCRdesktop code. its for
quick OCR Image and PDF files. this broke as well. So i decide to
reunion the code bases. I also did a lot of improvements and cleanups
while doing that.
you can now OCR files like images, PDFs (also with multiple pages) with
OCRdesktop. For this, its also quite optional now having a UI. because
OCRdesktop can pipe the result to STDOUT.
See here for the complete story:
there is also some goodie information about OCR the clipboard content in
my blog you might be interested in.
if you like my work you can support me on patreon if you want 8).
i have some other ideas i wanna bring in next couple of weeks:
- multiprocessing for OCR PDF files for performance speedup
- something essential as a loading bar while OCR process (lol)
- i have an idea for a while, tesseract creates some noise when OCR
widgets. i.e. a checkbox is detected as a Q. i want to suggest this
common pattern at details page.
i would say, we set then OCRdesktop on steroids 8-), i will return to
migrate Orca to an plugin driven architecture.
Am 15.01.22 um 23:12 schrieb Linux for blind general discussion:
> Hello Chrys,
> ah, great news! I consider this awesome project to be a fundamental and
> absolutely critical part of Linux accessibility. Unfortunately, despite
> high efforts to make software as accessible as possible, there will be
> still many inaccessible applications on all platforms, for quite a some
> Therefore, an OCR solution for these cases is an absolute must have for
> all serious blind users, and I'd like to thank you that you're keeping
> this alive on Linux, along all the other great work you're doing.
> OCRDesktop can be insanely handy, I've used it in many areas from
> accessibility of normal Linux apps, through installing software using
> wine to firing up virtual machines, which would be impossible without an
> equipment like this.
> So again, thank you!
> Keep up the great job!
> Best regards
> Dňa 12. 1. 2022 o 22:12 Linux for blind general discussion napísal(a):
>> Howdy List,
>> sorry for cross posting, but I thought you might be interested into that.
>> I just released OCRdesktop 3.0 with some very nice stuff :).
>> For those who don't know about OCRdesktop:
>> 1. i removed unmaintained "tesserwrap" bindings and use "pytesseract"
>> now. This is very good maintained and is able to use Tesseract 4.1.x and
>> Tesseract 5.0. Last didnt work with tesserwrap anymore cause by broken API.
>> 2. Removed all deprecation warnings, Modernize the code, use spaces
>> instead of Tab and all kind of cleanup goodness.
>> 3. i added an very nice new feature to analyze the color of an given
>> Text and its background. So you can figure what text is highlighted, has
>> the focus and all kind of useful stuff. its off by default and could be
>> used by "-O" flag as parameter. For this we need some new dependency
>> webcolor and scipy.
>> See the full story here:
>> Where to get:
>> - python3
>> - tesseract
>> - tesseract-lang-<yourLanguageCode>
>> - python3-pillow
>> - python-atspi
>> - python-pytesseract
>> - libwnck3
>> - GTK3
>> Optional Depencys
>> - python-scipy (for color detection)
>> - python-webcolors (for color detection)
>> I hope you have as much fun while using it as i had while working on it :).
>> have a good day and stay healthy in those crazy times.
>> cheers chrys
>> Blinux-list mailing list
>> Blinux-list at redhat.com
> Blinux-list mailing list
> Blinux-list at redhat.com
More information about the Blinux-list