Slashdot Mirror
Scanning Large Amounts of Pictures?
ClintJCL asks: "My wife & I are involved in scanning every photo I've ever taken in my life. She can lay down 4 or 5 pictures into the flatbed scanner at once, thereby saving the scanning time which is the bottleneck. But then she has to split them with Photoshop, which is also somewhat time-consuming. I've searched on the net for hours for a piece of software that would automatically split these 'Batch image scans' into single images and it just doesn't seem to exist. There are plenty of pieces of software to split a single image arbitrarily into sections for the purpose of loading faster on an HTML page (which I disagree with anyway and is not what I'm looking for). But -nothing- that seems to do any sort of edge-detection to determine what pictures exist in a given 'scan batch'. I'm out of resources. I've nowhere else to go. Perhaps someone can clue me in on a piece of software that can do this for me."
Finding the image edges from her bulk scans is one of the more trivial operations you can do on an image. Grab a handy-dandy image library for your chosen format (pnglib, jpeg, whatever) and write a couple pages of code and you're done. GPL it and help others with the same problem.
11*43+456^2
I was asked just yesterday to find something very much like this (except for some text processing and databasing thrown in). I haven't found anything out of the box that fits (not that I was smart enough to do an ask slashdot on it), so I think that I may have to code my own in the next week. Are you willing to pay a little cash for a custom solution?
It's been quite a while since I scanned anything, but I remember that with whatever scanner I was using at the time, in Photoshop you would do file->acquire->twain and it would bring up the scanning program. This program scanned the image lightly as a preview, and then let you select however many "jobs" you wanted at once, so you could select 1 square as job 1, another as job 2, another as job 3. Then it would make one pass and generate an image for each job.
As I said, this was a while ago and I don't remember the scanner, but it was probably some UMAX. The name "Mira scanner" stands out in my mind as the scanning software. You scanner may have this capability also; poke around a bit.
rooooar
This was my thinking. ImageMagick is well suited for the job. There is even a Win32 TWAIN API for Perl. Combined with the Perl API for Imagemagick you could even write that custom application that will automate the whole process.
Why not get a automatic document feeder for the scanner?
Why not just set up a Photoshop action to make 6 copies of the current image and set the action to crop each to an individual image. I don't know if you can script the name to save under automatically, but even if you can't you can just set up the actions to work on 10 or so original images and say to specific filenames and then just batch rename between scanning batches.
I have a higher end scanner (powerlook 3000) It allows you to do multiple scans in one pass.
ie you preview
you box all the pictures
you click scan
you can adjust each box color/exposure separate.
it scans each image as a separate file. Of course you have to preview each image which takes times.
You could write some software to do it. It might help to use a background matt of a consistant color though.
I think you really should consider scaning only images you care about and adjusting each one individually. If you really care about your images get a negative scanner. Scanning negatives is far far fat better than scanning prints.
Photoshop should be just fine for what you want to do.
There should be a pallete around that lets you create actions. You basically hit the record button and go through the steps you want to do. Map that to a keystroke and voila.
When selecting the picture, you could even used a fixed width box, so all of the pictures will have the same dimensions. For the action, you could have it copy what you've selected, create a new image and paste the selection into it. Then have it merge layers and auto-adjust the levels/contrast and then have it bring up the save dialog box.
All you're really left doing is scanning the set of pictures, clicking the mouse once (for the fixed width box), hitting the keystroke and then typing in the name of the file.
Karma: Totally excellent! (mostly affected by Bill and Ted and one Anonymous Coward)
Rufus is not an Anonymous Coward!!
- The border around every image shows a high contrast against the scanner background (which is usually white). This shouldn't be too much of a problem, unless you take lots of pictures of very light things.
- Your photographs are rectangular. This may sound silly, but it's a lot easier to find a rectangle than some arbitrary n-gon.
- Your photographs are placed so that edges of the photograph run parallel to the bed (i.e. you put the pictures down squarely)
If (1), (2) and (3) hold, then implementing this shouldn't be too bad --- I would use this algorithm:- take a scan of the background of the scanner (i.e. hit the scan button with no pictures on the bed) and remember this background image
- for each image in the input set
- perform edge detection
- use a Hough transform to detect lines in the edge map
- calculate the difference image D. If B[k](i,j) is the value of the k color band at pixel (i,j) in the background image and F[k](i,j) is the value of the k color band at pixel (i,j), then D(i,j) = max(abs(F[r](i,j)-B[r](i,j)),abs(F[g](i,j)-B[g](i
, j)),abs(F[b](i,j)-B[b](i,j))
- divide the difference image and the input image into rectangular regions using the lines that were detected with the Hough transform. For each region, calculate the average difference; if this value is large enough (i.e. over ~ 8-16), then consider that region to be a picture, otherwise it's blank space.
Intel's OpenCV library (look it up at SourceForge) can do most of the "tough" stuff for you (i.e. the edge detection and the Hough Transform). Hope this helpsOf course there are ways to do this on Windows. But it's hard to believe that no one has yet implemented it the unix way. There should be a utility where you type "picsplit *.jpg", and you get a directory of split files, sequentially numbered.
Programmers are always looking for projects, this sounds like a relatively easy one.
i dont understand how do i point my AOL to the free doctoral thesis channel.
ImageMagick comes with some command line utils that you can use to pull sections out of an image. Since you are scanning photos, I assume they are all the same size and have the same placement on the scanner. You could just write a script to grab 5 sections (or however many you are scanning) out of the image produced by the scanner. It won't automatically find the boundaries between each photo, but if you place them on the scanner carefully, you should be able to get consistent results.
The NetPBM package may have similar capabilities.
Need Free Juniper/NetScreen Support? JuniperForum
This is almost exactly what I was about to suggest, except that I would use string taped down at the edges (& pulled tight) instead of sticks. Why? It makes it easier to remove the pictures afterwards. You'll probably want to do the following:
Make a point to check that your strings haven't moved every hundred scans or so, and periodically check the cut images as well, just to save yourself from plunging ahead with a miscalibrated setup.
-- MarkusQ
Why don't you just scan the negatives? (that is if you still have them) It would be more economical than spending the next 2 years chopping up 8"x10"x150dpi.
Just a thought.
Dave
It is useful to have individual files to share with other people. But these might not be the same files you originally captured. You'd probably scan in the photos at the highest resolution and color depth your scanner can do. But when you attach an image to an email, you probably need to step it down a bit.
Anyway, you just have to have some kind of organizing software, if you've got so many photos that optimizing scanner time becomes an issue. If I were designing software that did this, I'd use a database, like the biolife example that comes with Delphi and Kylix. But now that I look at what's available, I see that most apps do store images as individual files.
Still, the principles the same. A good organizer can capture new images from your clipboard. So the procedure works out like this:
- Set organizer to automatically capture from clipboard.
- Place a bunch of images on the scanner.
- Scan. (This is the time-consuming part we need to minimize.)
- Drag a rectangle around an image, and click "copy".
- Repeat previous step for all images we just scanned.
- Go to step 2.
I guess that is a little less efficient than software that recognizes all the image boundaries. But the slight extra effort might not be as hard as finding software with the right kind of boundary recognition feature.First, scanning negatives with a film scanner will give you better results, but you state that you're poor, so that's out.
Second, you can split the image files from the scanner by size. This would require that you setup the photos that you scan groups of photos that are all the same size together, and get them placed in the same place on the scanner. Even then, you'd end up with either some white space around the photos, or they'd be cropped a tiny bit.
The third option (the one I think you had in mind) is image processing. pnmcrop (http://netpbm.sourceforge.net/doc/pnmcrop.html) will crop off the white or black border around a photo, but probably won't handle 4 photos at once. If you can fit your 4 photos within 4 static boxes, you can crop to those boxes by size, then use pnmcrop to chop off the extra white border.
Awesome furniture, accessories and cabinetry in Santa Rosa, CA: http://humanity-home.com/
this is an interesting problem, and since I've got GIMP installed and know a bit of Lisp (similiar to the Scheme that "Script-Fu" uses) I'm going to try and see what I can do using gimp's built-in functions. i'll set up a page at www.parseerror.com/split/ with my progress. i wouldn't gamble on when i'll have something worth using, but hey, you never know. i'll post back to this thread if i make any major break-throughs.
PLease mod up parent, it sounds like the best solution for a non-techie.
I figured an out of the box solution existed since most OCR software does this automatically when you scan page of mixed text and images, it will block out the page into sections to be OCR'd and images to be jpg compressed. I know Adobe Acrobat's scanning version is pretty good at this and let's you pre-define the default compression for the images.
Work for Change & GET PAID!
Every moment of my life the universe asks "Would you like to die now? It's free!" And the correct answer/reaction to give the universe is always NO.
Slow down, space cowboy!