[Insight-users] Re: On Finding Corners

[Luis Ibanez]

Hi Valli,

Instead of setting up the FloodFill iterator
by yoursefl,You may find more convenient to
use directly the ConfidenceConnectedImageFilter.
http://www.itk.org/Insight/Doxygen/html/classitk_1_1ConfidenceConnectedImageFilter.html

or the ConnectedThresholdImage Filter
http://www.itk.org/Insight/Doxygen/html/classitk_1_1ConnectedThresholdImageFilter.html

This filters use the FloodFill iterator internally
and define a connectivity criterion based on
the mean and variance (or a threshold) of the region
around one seed.

You could provide the image corner points as seeds
to this filter and run the filter in order to
recover the region connected to every corner.

You may have to run the filter four times. Once
with each image corner as seed.

Please let us know if you have further questions.


   Thanks


    Luis


=============================


Srivalli wrote:

> Dear Mr.Luis,
> 
>             Thankyou for your response.
> 
>  
> 
> In Insight/Applicatios I found example for 
> FloodFilledSpatialFunctionConditionalIterator. I have experimented with 
> it. In the example we have to set radius,center for spatial function and 
> seed point for FloodFillIterator. I have set the corners calculated 
> earlier as seed point. But, I have given some random values for radius 
> and center. I have given thresholded image as input for the iterator. I 
> didn't found any difference between the thresholded image and the output 
> of iterator.I didn't added connectivity to iterator.
> 
> I have some idea regarding connectivity. But,How to integrate 
> connectivity and floodfilliterator?
> Please give any suggestions.
> 
>  
> 
> Regards,
> 
> Valli
>