[Insight-users] a question about itkImagePCAShapeModelEstimator

[xiaolong li]

hi, luis

    i'm also working on a similar project, you said "that the images are
somehow
normalized, both in intensities and in the scaling and orientation
of the shapes contained inside of the images." i still have some questions
about this:

I tried to build a traing set for 2D image,   firstly i segement the
structure for all MRI images (say 10 samples ), the outputs are binary
images, then i  crop the images to smaller size which only contain
the structure and use them as input for PCAshapeEstimator, please correct me
if i'm wrong in those steps, but  why we still need intensity normalization
for binary images?

and,  how to perform a normalization of scaling and orientation on segmented
binary images (stucture) ?

for now i just crop the images in the same dimension (width and height), no
normalizations, i guess that's why i have some problem in segmentation when
using GeodesicPriorShapeLevelSetImageFilters,   i really appreciate
if you or other itk users can help me out.

thank you !

leo
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20081223/d7623e5f/attachment.htm>