__/ [ Carolyn Marenger ] on Sunday 12 March 2006 15:38 \__
> I have just downloaded 52 gif files that are scans of a 52 page document.
> They are quite clear, so I think they are suiteable for some OCR software.
>
> Are there any suggestions as to what software I can use for this? Is any
> of it included in the Power Pack edition of Mandriva LE 2005?
>
> I would really like to save my fingers the task of retypingthe fifty two
> pages.
>
> Thanks,
> Carolyn
This was discussed only yesterday in COLA. FWIW:
__/ [ x0054 ] on Sunday 12 March 2006 23:21 \__
> Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> wrote in
> news:dv0otu$1tbq$1@xxxxxxxxxxxxxxxxx:
>
>> __/ [ x0054 ] on Sunday 12 March 2006 09:06 \__
>>
> <cut>
>>>
>>> OCR - after that I can be totally happy :) That's all. And Photoshop
>>> runs grate under wine.
>>
>> It's already worked on. 15 developers are involved.
>>
>> http://sourceforge.net/projects/jocr
>>
>> Also see:
>>
>> http://software.newsforge.com/article.pl?sid=05/12/15/1848236&from=rss
>>
>> I used Photoshop for many years as a teenager. I have been using the
>> GIMP very happily for 4 years. Learn it thoroughly and you will see
>> that Photoshop becomes obsolete.
>>
>> Best wishes,
>>
>> Roy
>>
>
> I like GIMP, but Photoshop does not become obsolete, but GIMP is
> catching up quite well. I use GIMP quite a bit, just hate the interface,
> other then that it's perfect for me.
>
> As far as OCR there really isn't a good OCR software for linux at this
> point. I research it quite well, for a project for a friend, and the
> only production quality option available at the moment is ViviData, but
> it is ungodly expensive. JOCR isn't even close to production quality. I
> have tried others as well, no luck. But that's should soon be corrected
> because ScanSoft is working on a Linux copy of Omnipage, which is by far
> the best OCR software on the market right now. So I shell be totally
> happy soon :) Well, I'll also need a TTS client, don't like festivals
> TTS. But SAPI5 from M$ sucks just the same.
>
> BTW, another reason why M$ sucks. SAPI4 was written by a contractor,
> L&H, for M$ and sounds perfect, SAPI5 was rewritten by M$ from ground up
> and sounds horrible! Ignore the last bit if you are not a TTS geek.
>
>
> - Bogdan
|
|