Merge split image segments

Question

Dear Friends,
in an external PDF all images are splitted to smaller segments.
I heard this is quite common in PDF generation due Postscript.
I am placing this external PDF on an Indesign document page which is also exported again as a final PDF/SWF/JPEG.
But in the final PDF/SWF/JPEG there are white lines visible in the images, when they appear, allways between two image segments.
I heard this is due to rounding errors of position x-/y-coordinates of these image segments rendering them not perfectly onto their intended position.
I think about two general solutions:
Merge all image segments to a complete image.
Being able to position these segments with lock / automatically to make them perfectly edge to each other (but what would be with these rounding errors?).
This is a quite common problem to designers who want to further use the ressources or pages of an external PDF.
Often the senders of those external PDFs have not even got the original data (careless/lost/etc.).
It would be great if there is a reliable, solid solution to this problem.
Thank you for answers in advance!
With best regards
Inno

Answer 1

A naive (but possibly sufficient) implementation could just be to keep a table mapping (x,y) pairs to countries. To keep the table smaller, you could map 10x10 blocks of pixels to countries.
Ultimately this is about storing and then managing data.
One hack (in my opinion) is to check the color of the clicked pixel in the image of the map. But that is nasty for a variety of reasons (such as, it keeps you from being able to easily change the map image, and things like text in the map (like the names of countries) kind of screws that up. But it's an option.
You can also keep lists of pixels in sequence describing a polygon, and then use certain algorithms (which I don't recall right now) to determine whether a given point is inside or outside a given polygon.
Also you can do things like "binary space partitions", I think.
These last two things are computer graphics techniques, and are non-trivial to implement. I dimly recall them from college but recall no details. You can get a computer graphics textbook (which tend to be expensive) to get more info, or these days there's probably a wiki about it somewhere.
But I'd suggest keeping it modular. Add a layer of abstraction, writing an API around this functionality. Then write an implementation with a simple table of locations or blocks of locations, and if this proves to be insufficient later, swap it out with something more sophisticated.