aligning two lists of points in 2D space

Hi,

I have two lists, L1 and L2, of points (x,y) in 2D space. They represents location of 'events'. Those two lists are misalign (mainly translation, could be rotation). Each list, display clustering, but shape of a cluster in each group is different, let say one is filament other blob. I tried to use different image aligning alghoritms (i could bin my lists into an image), but they fail. when I overly them I see clear translation, blobs should be near filaments, but i would like to automitize it. Most importantly, two lists do not represent the same structures, which were just misaligned. Do you know of any alghrotim, which could help me? thanks in advance,

Ina

Answer 1

[Matt J]

I don't know exactly what you mean by "two lists dont represent same structures", but have a look at

http://www.mathworks.com/matlabcentral/fileexchange/26186-absolute-orientation-horn-s-method

[topolowa]

Hi, thanks will have a look on that. To clarify what i said in my question: list of points can be binned into an image; structures in an image could be any of shape. Saying that I wanted to stress, that I try to align two images, which are not the same. You could think about that as two image chanels from the microscope. Each chanel labels different e.g. cellular structure, which shape is different from the other. There is lots of algoritms an tools in matlab, which align or register two images, which are actually images of the same thing, but transformed in some way. My images are different, but they are well 'colocalized'. Alternativelly, I am asking for alghoritm to align two list of 2D points, because that is how my original data looks.

thank you very much

ina

[topolowa]

Just to add, my lists of points have differrent numbers of points, e.g. Mx2 and Nx2, N~=M