[Insight-users] RecursiveGaussianImageFilter

[Luis Ibanez]

Hi Silvano,

You are right, the results of the derivative shouldn't
change when you add a constant value to the input image.

What is the pixel type of the image that you are using ?

Are you performing the smoothing on the other directions ?
Note that for computing the first derivative along X
you still need to do smoothing along Y and along Z.

Could you please set up a minimal example where you observe
this behavior and post it to the list ?


Thanks


   Luis


------------------------
Silvano Agliozzo wrote:

> Hi to everyones,
> 
> I implemented an application for the calculation of first and second order
> partial derivatives of a 3D image by using the
> RecursiveGaussianImageFilter. I observed a strange behaviour by adding a
> large costant (positive and negative) to the values of the 3D image.  
> Actually the results are completely different each other and with respect
> to the case without the addition of the costant. This does not happen by
> using the DerivativeImageFilter, but it is unfortunately slower. From 
> the theory, adding a constant to a function does not change the value of 
> its derivativesB.
> 
> I do not understand what's wrong in this filter or with my implementation.
> Any suggestions ?
> 
> thank you,
> 
> Silvano
> 
> _______________________________________________
> Insight-users mailing list
> Insight-users at itk.org
> http://www.itk.org/mailman/listinfo/insight-users
>