[Insight-users] [Insight-developers] down sampling

[kevin cavin]

Hi, Luis: 
 
   Thanks for your  reply.  I attach the result for 3D registration for orginal resolution of images and the sub sample version.  The sub sample image factor is 2 for x and Y, and 1 for Z. It means that the each slice is half copmared with the original slice. But the number of slices is not changed. 
 
    Regarding your question about what components are you using ?

  1) Image modality ?
  2) Transform ?
  3) Optimizer ?
  4) Metric ?
  5) Interpolator ?

I attach part of my code in python 
  

registration = itk.MultiResolutionImageRegistrationMethod[itk.Image[itk.F,3],itk.Image[itk.F,3]].New()
imageMetric  = itk.MattesMutualInformationImageToImageMetric[itk.Image
[itk.F,3],itk.Image[itk.F,3]].New()


transform    = itk.VersorRigid3DTransform.D.New()
finalTransform= itk.VersorRigid3DTransform.D.New()
optimizer    = itk.RegularStepGradientDescentOptimizer.New()
interpolator = itk.LinearInterpolateImageFunction[itk.Image[itk.F,3],itk.D].New() 

 
 I also attach the result repsectively using the orginal volume and sub sample volume 
 
 for orginal volume I get reasonably registered result. the result is 
 
Final Registration Parameters 
versor X      =  -0.001937
versor Y      =  0.003333
versor Z      =  -0.002403
Translation X =  -0.250085
Translation Y =  -4.207276
Translation Z =  -782.659578
Iterations    =  45.000000
Metric value  = -0.124261
 
for sub sample volume the registered result is empty, which means only defalut pixel values
Final Registration Parameters 
versor X      =  0.000695
versor Y      =  -0.011173
versor Z      =  -0.000622
Translation X =  -0.034234
Translation Y =  -2.751356
Translation Z =  -794.621228
Iterations    =  45.000000
Metric value  = -0.147362

   So I am wondering where goes wrong?  Please let me know if there is anything unclear. Thx 
 
Wen 
 


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20090714/3622f9d4/attachment.htm>