On Mon, 9 Feb 2004 23:17:32 -0000, Witchy <[EMAIL PROTECTED]> wrote:
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Phoebus R. Dokos (ÃÃÃÃÃÃ Ã. ÃÃÃÃÃÃ) Sent: 09 February 2003 20:59 To: [EMAIL PROTECTED] Subject: Re: [ql-users] Thor manual
...And the page nos are rather mixed up so I am converting to text page by page instead of the whole lot :-) Good thing that my OCR learns fast :-)
Hi Phoebus,
What software are you using? I'm trying to OCR bits of the Complete Spectrum
ROM Disassembly using Caere OmniPage Pro and it's slow going to say the
least.....
Cheers
I am using SimpleOCR which started off badly but learned after 5 or 6 pages despite the very bad scan. I had to import the PDF to Photoshop and process every page individually first before I could do the thing. Unfortunately that left me with 54 files that are a pain in the you-know-what to reassemble :-)
Omnipage is very good (it's no Kurzweil or Recognita but still) if your originals are good. Best results are obtained with a good contrast, reduced noise (ie photocopy speckles) and properly oriented pages :-) (Same as most OCRs that is)
Phoebus
