[Insight-users] question on itkGaussianDerivativeOperator

[niccolo camarlinghi]

Dear ITK users,

I'm trying to use the version of  itkGaussianDerivativeOperator included in
ITK3.20.
Browsing the source and trying to use the class, I noticed that setting the
flag SetNormalizeAcrossScale to true doesn't affect the 0 order derivative,
i.e. the kernel  of 0 zero order is the same with or without
 SetNormalizeAcrossScalet setted to true.
In my knowledge using scale space normalization means that the Gaussian (0
order derivative), used for the convolution, is not normalized to 1 as
usual, but its normalization is equal  to sigma^n, where sigma is the std
deviation of the gaussian and n is  the number of dimensions of the kernel.
Is this correct?
I'll be very grateful if someone can explain to me this reason of this
implementation.


Thank you and best regards

Niccolo' Camarlinghi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.itk.org/pipermail/insight-users/attachments/20110324/40ca2119/attachment.htm>