Lee writes: > At 13:36 11/06/2002, Leon Brocard wrote: > >ken sent the following bits through the ether: > > > > > I have two image, A is a photo, B is a part of A. How can I know > > > where (x,y) is the photo B in Photo A? > > > >If B is an exact partial image of A then look at the images like a > >string and look for the right seqeuences of colours. > > Which will work if B is aligned neatly, like a nice clip of A. > > If it's not aligned, then you're probably going to be stuck for > a while. Artificial Neural Networks could easily find B if it > really is a subset of A: there are lots and lots of papers on > the matter around.... here's my ANN reading list, if it helps:
This problem is called "image registration" or "image stitching" or "image mosaicing". There are plently of papers written on the subject, most of which don't use AI techniques. The most widely used method is getting the Fourier transform of both images and comparing them; but this only works if there is no change in rotation between the images. I, personally, have written a paper to do just that using Zernike moments. Many other approaches exist, each with it's ups and downs. > >A more > >interesting case is covered by the following paper (interesting links > >too): http://citeseer.nj.nec.com/336979.html > > But it looks like both our suggestions would require something > more powerful than little ol' Perl (no flames, please). You certainly *CAN* do it with Perl. The question is whether you *WANT* to! --Ala
