I'm trying to move all (or as much as possible) my work from Windows to Linux. I scan photo archives to make digital archives of old photos. My windows program scans as many photos as I can fit in the scanner at once and makes individual files out of them.

I'm new to linux and so far all I have tried using is kooka, but I can't see how it could scan multiple images together as separate files at once.

Anyone know a more versatile software for scanning?

have you checked freshmeat.net?

http://freshmeat.net

Also what scanner software are you using that does this?
That sounds really cool and unlike anything I've ever used

I'll have a look at freshmeat, but if it doesn't state what I need in the description it will be impractical for me to download because we have EXTREMELY slow internet connections (I'm in India right now). That's why I'd like to know if someone has used a software with this functionality.

The scanning software I use in windows is Scan Wizard v.5, which came on the CD with the Microtek ScanMaker 3840 we have here. When you use it in advanced (or expert) mode, then you have a lot more direct control over the scan job, including the ability to identify multiple images and scan them with one pass and end up with a separate file for each. I love it!

What I would love even more is to do this job under the Linux side of my computer...

Can't find anything on freshmeat that even remotely looks like what I need.

Not much help but look at this for maybe an idea. If using Sane and Gimp the in sane options need changed to do auto document scanning.
http://solutions.brother.com/linux/s...nux_faq.html#2

I think one can use gimp or similiar program on the command line and written a small script to do batch scanning and saving. I am sure someone knows more about this idea. I will give this some more thought and post if I come up with and idea.

Quick question, does the scanner work in linux now?

Brian1
" Google the Linux way @ http://www.google.com/linux "

Thanks for your thoughts. I don't think those brother related suggestions can work because my scanner is just a regular flat-bed style scanner. It's just that I'm putting 2-5 photos on the glass (depending on their size) and in Windows I can scan the whole thing with one pass and each photo gets saved with it's own sequential file name in the directory I specify.

Yes, my scanner is working in Linux. I can scan one photo at a time in kooka, but I don't really use it. I always go back to Windows because I can get so much more done in so much less time.

sounds like omnipage or the HP stuff which was ripped off from caere....



the term you are looking for is AUTO-REGION.....

when a a4 sheet is scanned it is looked at by software to recognize wtf is an image and whats text so that images are not fed into the OCR engine....

id say that this is not available via linux at the moment because the source of the original app is very closed source....

identifying images as opposed to text is pretty easy as images have shittins of data in such a small area and also color differences etc....

things to realise about the HP/Caere apps........

auto straightening pages, Splitting them into text/image areas saving them off.....all possible manullky in linux using gimp etc.....if you build a frame on your flatbed which isolates the locations of photos and then work out the values once scanned into Gimp or something you can create macros to auso cut paste to new and save off the file.....

I actually do this via Photoshop in Wine as i am a adobe born artist.....

Did a little browsing. Have you tried scanimage. If not and if I understand it try the following:

type ' scanimage -L ' to list scan devices.
With that if then try ' scanimage -d hp:/dev/usbscanner0 --format tiff --batch '
Change -d ***** to match what ' scanimage -L ' reported.

For more info check your man pages. ' man scanimage '
There also appears to be a gui interface but could not find much on it.

I cannot remember right now but there is a command that can convert one image format to another format. I need to look through my old notes or find it on the web.

See how this goes for you and get back to me.
Brian1

thats probably useable to some extent for what he wants....



if scanimage has the ability to target XY coords or something similar you can make a frame and run a script to call scan image 5 times on small sections X&Y and pump them out as a 5 different files with pic 1-5 named etc...

i would hope the line would look something like


scanimage -d /dev/HPSCANNER -f /home/mypics/PROmptforname -xstart 100 -xfin 200 -ystart 100 -yfin 100 -format gif......

that would work well enough....script that 5 times with a filename prompt and some sort of confirmation once each line ran would probably be needed to trigger off the next scan command as the scanner needs to reinitialize after each scan so timing may be a problem on the script...