[nflug] Stand-alone scanner?

[Sam Stern]

>-----Original Message-----
>From: Joe [mailto:josephj at main.nc.us] 
>Sent: Sunday, April 30, 2006 04:04 AM
>To: samstern at samstern.net; nflug at nflug.org
>Subject: Re: [nflug] Stand-alone scanner?
>
>Cool. You are amazing. This was way too simple to think of. My scanner 
>is super old, and only works on Win 98, so I haven't done any 
>ocr in Linux.

Hmm. Scanner support in Linux has come a long way; HP and Epson have made
some really good progress with getting most of their scanners working under
sane. My HP 2500's scanner and printer work better in Linux than in windows!
So, what is the make / model of scanner you own? Perhaps we can find the
right magic to make it work.

>Assuming I save the pictures of pages .... I have a couple of cheap 
>digital cameras that save as jpegs.
>
>1) Do the new, better cameras save as tiff (as an option)? 
>(What format 
>do I want them in - pnm?)

Yes, the better cameras save in jpg, TIFF and RAW. The best formats for
scanning for OCR are TIFF or RAW. Also, I would expect a no compressed
format like bmp, gif or png to work better than jpeg. I do not know how well
a jpg works but a 100% jpg is probably not the best format to be using. You
should be able to check the dpi by looking at the jpg exif tags of the
image. I do not know what application one would use on Linux to check these
tags. I use ifranview in windows to get this information from images.


>I can probably convert jpeg to tiff (or pnm, see below), but an extra 
>conversion would probably degrade the image.
>

I doubt that would actually help.

>2) What software do I use on Linux to run a tiff file through ocr 
>conversion?
>I found kooka, but it seems to start at Mandrake 9.2, so maybe it will 
>run on my next pc. Ditto for ocrad ad, which it uses (10.0+)
>

I'm afraid that desktop linux is not my strong suit. I'm more of a server
and security person. Hopefully another member of the form can help. That
said, I would check to see if programs like xv, kview, gimp would help. 

>I installed gocr, but haven't tried it yet. It uses file 
>formats I'm not 
>familiar with. I'm not sure how well developed it is. I 
>installed djpeg 
>and cjpeg as shown in the examples. I also found jpegtopnm (based on 
>djpeg) already on my system.
>

Sorry, I wish I knew more about desktop Linux; right now my main desktop
apps have no effective analog in the Linux world (MS Access, PDF Form
creation, Real Arcade, etc).

 I have access to a Suse 10 system that I keep updated with apt to the
latest kde. Unfortunately, all I do in the gui as far as desktop Linux goes
is some light email, web browsing, Word processing, Spreadsheet and the
occasional mahjong game. I usually use the gui for controlling vmware
server, workstation and remote access when my other windows systems are
otherwise occupied in pen tests or playing a game.

>Tomorrow, I'll boot into windoze and see if I have anything already 
>scanned that I can test with. If not, I'll create something.

Good luck. If you need some win98 help; email me off list I know allot about
Win98 (and integrating it with linux). 

>
>Joe
>

_______________________________________________
nflug mailing list
nflug at nflug.org
http://www.nflug.org/mailman/listinfo/nflug