[Insight-users] using filters on sample data

[Aditya Chandramouli]

Hi Luis,

Basically, I'm trying to implement an Expectation-Maximization routine for an 
image segmentation. 

=> I start by iterating over my image and calculate a Gaussian mixture(with 
the help of GaussianDensityFunction in the Statistics package).

=> I store the results in a ListSample with measurement vector type 
itk::Vector<float, NumberOfChannels>

=> The next step of my implementation requires me to process these results and 
apply a simple Low Pass filter. This is where the problem lies.

=> After this, the filtered data has to be further processed and stored back 
to an image.

Hope this is clear.

Thanks a lot for your patience in looking into this.

-aditya

PS: To try to get around this problem, I've so far tried to create an empty 
image and iterate over it, filling the pixels with the data that I would 
filter subsequently. However this means I have to use the inefficient Set() 
method to set the pixel values. Apart from that, I'm getting a segmentation 
fault (presumably because the image is empty to begin with)...


On Sunday 30 April 2006 01:38, you wrote:
> What kind of data is exactly stored in your ListSample ?
>
> do you have pixel values ?
>
> or coordinates ?
>
> How did you got your ListSample ?
>
>
>   Please let us know,
>
>
>
>     Thanks
>
>
>        Luis