[Insight-users] compute the deviation of a certain region

[Xi Liang]

Hi all,

I am trying to mapping an image using sigmoid function to enhance some
bright region and remove noise in dark regions.  There are two parameters
one is beta and alpha, where intensity below beta - alpha is suppressed to
the minimum value of the image, and intensity above beta + alpha is mapped
to the maximum value.

In order to find a proper value for beta and alpha, I use kmeans (e.g. k =
3) method to find a highest centroid C_3 and set beta = C_3. This has been
done without problem. Im trying to find a proper alpha now. I think the
deviation of the cluster with the highest centroid value C_3 would be a good
start point.

The problem I have now is how to compute the deviation of voxels that fall
in a certain cluster. It is easy to write an iterator to compute this. I
would like to know whether there is any other method to do this. I am
thinking the following possibilities:

1: using StatisticsImageFilter, but there is no function to set a mask or
region.
2. maybe using a spatial structure, instead of image would be better?

Cheers,

Xi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20101104/cb0bb4b5/attachment.htm>