Automated photo scanner?

[xennex81]

I can't write everything in one post because I don't know when I'll be banned again. So that is perhaps the most important reason, I don't know. Also I don't stand to lose any text this way or run into people posting between me after which I have to change anything. Sorry about that.

Also, to drop some hints, I feel the Apache Jira system (it is a commercial package but licensed for free to open source like ASF) is a very good system to collaborate on with people. It offers not only bug reports but also improvements or patches and suggestions, mostly anything you would like to tell a developer or a development team can be featured and used in that Jira. Ant is a bitch but a good one, Maven is also a bitch but a slightly worse one I think, maybe that makes it a better bitch, I don't know.

Documentation on Maven is thoroughly bad <-- that is the kind of hate you think I have? It is justified -- > but that on Ant is reasonably good and I kinda know....

Well let's say I hate writing Unit tests. Signing out.

And oh before you say so, it was DoYouKubuntu who made this thread only about this and then you Qqmike exagerated that and made it even more about me or my endeavours. She had the chance to reply to my constructive answers on her subject matter but she chose not to and to ignore it completely and only talk about her hate of Windows (mostly).

[DoYouKubuntu]

DoYouKubuntu is a valued and well-liked member here, and IMO asked a perfectly reasonable, valid question. In fact, I am currently in the same situation, having to scan a bunch of old photos and organize them.

As usual, xennex81, you have done what you usually do in your 363 posts to date (though it would be difficult to do an ex poste analysis of your posts because many have been deleted, including deleted by you or edited by you):

(1) You have made it all about you, your precious self;
and
(2) Made it clear how bad Linux is, in your opinion.

Furthermore, you go on with some psycho-babble, analyzing DYK's motives and behavior. Two problems with that:

(1) It is inappropriate, in response to a technical issue DYK raised.
and
(2) Who are you to analyze anyone?


DYK makes a good point: Why should anyone have to defend Linux on a Linux forum?

You have repeatedly made your position crystal clear: You dislike--hate?--Linux; you have written many dozens of page-equivalents explaining why. What exactly is your purpose here? Why are you here? We certainly don't need this, but what are you getting out of it, except a theater in which to vent and pontificate?


DYK can defend herself. I'm simply making the pattern of your responses clear.


---------
Quotes from xennex81, Post #8.

Thank you, Qq. Really, your words mean a lot to me.

[Qqmike]

(an elephant faithful ... 100 percent ... -- do you recognize that?)

[GreyGeek]

Thank you, Qq. Really, your words mean a lot to me.

This forum has been and always is about support for Linux, Kubuntu and the people who use it, and Windows on ocassions.
Qqmike speaks for all of us, DYK. I have not read a single encouraging or informative post by xennex81. I doubt he will be posting here anymore.

[vinnywright]

WoW @xennex81 hasent been back a few days and managed to get banned again ,,,,,,,,, well we can do without the psychobabble ,,,,especially when attacking a valued member of our community like @DoYouKubuntu

I'm usually not for banning ,,,,,,,,but enough is enough

VINNY

[DoYouKubuntu]

This forum has been and always is about support for Linux, Kubuntu and the people who use it, and Windows on ocassions.

Indeed. And it's generally a warm, welcoming place for support.

Qqmike speaks for all of us, DYK.

Thank you.

I have not read a single encouraging or informative post by xennex81. I doubt he will be posting here anymore.

It's his loss. This is such a great, tight-knit place.

[DoYouKubuntu]

WoW @xennex81 hasent been back a few days and managed to get banned again ,,,,,,,,, well we can do without the psychobabble ,,,,especially when attacking a valued member of our community like @DoYouKubuntu

I'm usually not for banning ,,,,,,,,but enough is enough

Thank you, Vinny.

[GreyGeek]

...

Can anyone recommend a scanner, whether it's specifically for photos or not, that can automate this process? I want to stick a bunch of photos in a hopper/tray and have the scanner feed them in one at a time, then just name them sequentially [or whatever] with some generic name I start with, saving them all in the same location. Does such a creature exist? FOR LINUX? I don't do windows, so there's no point discussing any of those.

Of course you'll need a scanner with a document feeder. Like all feeders the rails on the feed tray have to be adjusted for the sized to the stack of photos you are feeding it so none will twist and miss-scan or jam. Secondly, the scanner software, which Linux has for HP and other printers, will automatically create serial names for the photos. There-in is your first problem. What is on "000627.jpg"? For that you'll need to create a database which uses an integer as a primary key and at least two fields: filename and description. Perhaps a tag field also. Then you'll need a script which reads the image files in and saves them to a database record using the integer value of the file names. After that you'll need a GUI app which reads each record in, displays the image and presents a text block into which you enter a description, and a tag field if you use one. Dating the images is another issue. The scanner won't read image meta data (date, size, type, etc...), so that will be a manual process.

Image sizes will depend on the resolution you use. BW will use fewer bytes per image, but you will have to scan BW's and color separately. With sizes ranging between 100Kb and 10Mb, "thousands" of images could total 100Gb. And, double that to back up your treasure trove of pictures after you scan them, describe them and tag them. Oh, you will want to index every field, so on the safe side double the HD size requirements.

PS- Several years ago I used an HP 5610 to read a couple hundred photos. The problem was that the ADF (auto doc feeder) was not designed to accept heavy stock or photos, just paper. Also, photo sizes varied so much I ended up placing several photos on the scanner and then using Gimp to pick the individual images out of the big single image.

[DoYouKubuntu]

Of course you'll need a scanner with a document feeder. Like all feeders the rails on the feed tray have to be adjusted for the sized to the stack of photos you are feeding it so none will twist and miss-scan or jam.

I wonder if there's something I can add, something external, an attachment of some sort, to my HP Deskjet F4140 All-in-One? Or if there's some way to use it to do what I want?

Secondly, the scanner software, which Linux has for HP and other printers, will automatically create serial names for the photos. There-in is your first problem. What is on "000627.jpg"? For that you'll need to create a database which uses an integer as a primary key and at least two fields: filename and description. Perhaps a tag field also. Then you'll need a script which reads the image files in and saves them to a database record using the integer value of the file names. After that you'll need a GUI app which reads each record in, displays the image and presents a text block into which you enter a description, and a tag field if you use one. Dating the images is another issue. The scanner won't read image meta data (date, size, type, etc...), so that will be a manual process.

Honestly, I'm not seeing this as an issue. Not an immediate issue, anyway. My objective, first and foremost, is to get the photos scanned/saved to a hard drive. I'll do them in groups based on...some criteria. Perhaps a time period, like family photos from the 1930s, or perhaps a theme, like travel photos or baby pictures. So they'll already be roughly sorted by the time I get around to fine-tuning their names. To start I'd just want something generic, like "1930_family_XXXX.jpg" or "Disneyland_XXXX.jpg", etc., where "XXXX" is incremented each time another photo is scanned/saved.

Image sizes will depend on the resolution you use. BW will use fewer bytes per image, but you will have to scan BW's and color separately. With sizes ranging between 100Kb and 10Mb, "thousands" of images could total 100Gb. And, double that to back up your treasure trove of pictures after you scan them, describe them and tag them. Oh, you will want to index every field, so on the safe side double the HD size requirements.

Oh, yes! This will DEFINITELY be a space hog! Thank goodness for multi-terabyte drives, right?

PS- Several years ago I used an HP 5610 to read a couple hundred photos. The problem was that the ADF (auto doc feeder) was not designed to accept heavy stock or photos, just paper. Also, photo sizes varied so much I ended up placing several photos on the scanner and then using Gimp to pick the individual images out of the big single image.

I've THOUGHT about doing what you did with the GIMP. But I really...really...don't want to!

[GreyGeek]

The HP F4140 is supposed to have an ADF. Your's doesn't?

[DoYouKubuntu]

The HP F4140 is supposed to have an ADF. Yours doesn't?

Its document feeder has only ever been used--in my house--for loading 8.5" x 11" computer paper into the printer. I know that it's adjustable, to a degree, although I don't know (and am too lazy to go check) if it can be adjusted down small enough for old, small photos. Besides, even if it could, what software would I use to do the actual process I want?

[oshunluvr]

Out of curiosity, I walked over to my new-ish HP 8600 Pro Deluxe and placed a roughly 5"x4" card in the document feeder, selected scan, 600 dpi (the max allowed for ADF scans), and color JPEG, put a USB drive in the printer (has mem card slots too), and hit go. Result was a 5088x3045 RGB jpeg of respectable value, but it was turned sideways a bit as it was too narrow for the ADF guide.

Software: I have successfully used my ADF with Xsane and multiple pages also. I haven't tried multiple photos yet, but it appears doable IF they are large enough for the ADF. Simply selecting ADF and the number of documents to scan does the trick. You can select the filename and number sequence to use at scan time.

I think the challenge will be the feeder quality. I suspect with most if not all ADFs you will have to separate your photos by size - width at least. Plus, at least mine, wouldn't adjust small enough to scan the littler photos well. When you add in the possibility of "fancy" stuff like scalloped or "pinked" photo edges and left over glue and paper on the back of the old photos removed from albums, the likelihood of ADF failure goes way up.

My conclusion: 1000's of photos? Really? I'd suggest offering a stack of photos at a time to your ADF and Xsane. Compare them after they are scanned to the image and separate them into two piles: done and not-done. Delete the bad scans and repeat. At least you will have (hopefully) diminished the workload to some degree. Once you have a pile of not-done photos you'll have to decide whether or not to invest money in more hardware or go the flat bed route. You might consider a few evenings sorting them in advance of scanning - tossing repeats or unneeded near duplicates. How many pictures of Aunt Bessie at the picnic does one need anyway?

BTW everyone; Who calls people cow turds anyway? I've had that dude on "Ignore" for months now, but took him off to read this thread. All I can say is WOW...

[DoYouKubuntu]

Thanks for the input, oshunluvr, and especially for doing some experimenting! Your results help put things in perspective, and give me a bit of insight into what to expect. You're right about left over glue/adhesive, scalloped edges, and other peculiarities that I'll encounter when scanning old photos. But since this project is already years...decades...overdue, I'm resigned to it taking time, i.e., I don't expect to do it in one day. Good thing, too, since it'll probably take weeks--at least!

BTW everyone; Who calls people cow turds anyway? I've had that dude on "Ignore" for months now, but took him off to read this thread. All I can say is WOW...

Yep!

[GreyGeek]

Just wondering ... In the long term would it be easier to take your photos to a business which scans photos in bulk and gives you the results on an HD or SSD?

more later

[DoYouKubuntu]

Just wondering ... In the long term would it be easier to take your photos to a business which scans photos in bulk and gives you the results on an HD or SSD?

more later

You know what? I was actually thinking about that. Let someone else deal with it! Whatever cost is involved would probably be worth the savings in time and aggravation. I guess I'm just hardwired to do stuff myself...but this might be a great example of an exception to that rule.