[Insight-users] Re: model-to-image fitting for gaussian images

[Luis Ibanez]

Hi Ali,


1) By "Gaussian images" do you mean that the image has small sub regions
    where the spatial distribution of pixel intensities is Gaussian ?

2) Is there anything else in the image, in addition to these Gaussians ?

3) How diverse are the Gaussians ? do you have a range of sigmas that
    spans more than an order of magnitude ?

4) Are the Gaussians radially symmetric ?
    or do they form ellipsoids

5) What is the dimension of your image, 2D ?  3D ?

--

If the Gaussians are not very diverse, you could use a FFT transform
and filter them in Fourier space. That will be equivalent to doing
convolution in physical space. That will give you peaks for solving
the location problem. (e.g. finding where the Gaussians are).

 From those locations you could estimate the sigma of each Gaussian
by integrating its intensity values from the top of the peak, down
to pixels with intensities > than peak / 3.0.  If you divide that
integral by the peak value, you will have a number that can be
correlated the sigma value. This has the advantage that you don't
have to "fit" a gaussian model to the small intensity patch.


   Regards,


      Luis



-------------
Ali - wrote:
> Anyone?
> 
>  >Hi,
>  >
>  >Assume an image which contains a set of randomly distributed gaussian 
> images with diameters of 3 - 10 pixels. The aim is to:
>  >
>  >(1) locate the gaussian images
>  >
>  >(2) fit a guassian model to these images
>  >
>  >The model-to-image registration approach, as described in itk software 
> guide, seems to be a good way of implementing the 2nd part, is that 
> right? Do I need to write a metric myself, or is there anything 
> out-of-the-box currently in itk?
> 
> ------------------------------------------------------------------------
> Are you the Quizmaster? Play BrainBattle with a friend now! 
> <http://specials.uk.msn.com/brainbattle>
> 
> 
> ------------------------------------------------------------------------
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users