Take the 2-minute tour ×
Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

I have a set of images, which are parts of a bigger image.

I would like the images to assembled as automated as possible, to recompose the whole picture.

How can I do that?

Please note that this is not pictures. I've tried Hugin and Autostitch, but here, I have a flat image and the softwares didn' worked as expected.

I have nearly 10x10 small images. All images have the same size, but parts of the bigger can appears on two pictures, that my overlapping a bit.

[Edit] To illustrate, my wish, here is a very simple example with a 2*2 picture :

enter image description here

I have 4 images (borders are not in the images, its for comprehension). The solution may detect automatically how to position each images, and produces the final image

[Edit2] A bit of background: the process, whatever tool it implies, is targeted to people that have low level of image editing skills (so I am!). Using some basic tools like autostitch is ok (if it works, which is not the case for autostitch). And most of all, paid software is not an option in my context. Using photoshop is then prohibited.

The process will be executed several times, by several users. The acceptable amount of time for the whole manual process (excluding potential computing time then) should be less than 10 minutes.

share|improve this question
    
So your sources aren't in a particular order? –  slhck Nov 14 '12 at 14:17
    
unfortunately not. In fact, fragment can be any part of the picture, I said 10*10, but fragments are not aligned, and there's no guarantee that there is exactly 10 pictures per row, or 10 picture per column. I can afford roughly positioning images myself, until I don't have to position at pixel precision. The solution may recognize the picture and stick automatically when pixels match. –  Steve B Nov 14 '12 at 14:20
    
If it can help, the picture fragments have the same "zoom" level. –  Steve B Nov 14 '12 at 14:21
add comment

1 Answer

Do you have a program like GIMP, which allows you to create layers, and adjust their transparency?

Do you have the time to manually assemble this picture yourself, or is this some kind of academic exercise where you want to find the software that will automatically assemble for you?

Because the first thing that came to my mind, was importing each picture fragment into a Photoshop image. The first fragment you import you leave solid. Each subsequent image you import, you put on a new layer and make transparent... and move around until it lines up.

Unfortunately...

I would like to assemble the images to recompose the whole picture.

and

The solution may detect automatically how to position each images, and produces the final image

seem to be in conflict to a degree. The first implies you want to do this. The second implies you want software to do this for you. My solution satisfies the conditions of the first statement, but not the second.

EDIT - I removed the reference to Adobe Photoshop, which costs money, and replaced with GIMP (as well as a link to it) which is free.

share|improve this answer
    
I've rephrased a bit my question to be more clear. The idea is to automate the maximum of the process. I've also added a bit of requirements. Your solution then is not satisfying, especially at the point and move around until it lines up. This is exactly what I want to avoid, because it will become quickly a huge amount of work. Even paint may do the job otherwise I think –  Steve B Nov 14 '12 at 14:52
    
you must not have all that much experience with using transparent layers. If you believe Paint would do the job, then what you are imagining is using the EDGES of each fragment to attempt to line up the fragments. However, if each fragment is a transparent layer, you are using the whole fragment to line up with the one below it. Makes the assembly process very fast for the first second for rough lining up. Think of it as lining up tracing paper on a light board. –  Bon Gart Nov 14 '12 at 15:21
    
Ok, I can see your point. It will be faster using layer with transparency. But it will still be a manual process to make this exactly match. –  Steve B Nov 14 '12 at 15:49
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.