[Insight-users] Re: Running examples in the SoftwareGuide

[Luis Ibanez]

Hi Jianmin,

--------------
jianming.liang@computer.org wrote:
> Dear Luis,
> 
> Thank you very much for your kind instructions.
> 
> 1: Yes, now I can reproduce the results of
> 
>      FastMarchingImageFilter.exe
>      ShapeDetectionLevelSetFilter.exe
> 
>    in ITK/ITK1.2bin/InsightToolkit-1.2.0/bin/Debug. If I may, I would
>    like to suggest to include the parameters for StoppingValue in the
>    table on page 247 in User's Guide.
>



That's a good suggestion,
The stopping value has now been added.
(a bug in the counting of  argc parameters was also fixed).




> 2: I have also tried
> 
>      GeodesicActiveContourImageFilter.exe
> 
>    in the same directory. I guess that you swapped the seed points in
>    the table on page 259. Even if I swapped them back, I cannot
>    reproduce the result for white matter. Would you please check the
>    parameters. I should note that I changed the parameter in an
>    anisotropic diffusion filter for smoothing from 0.25 to 0.125 in
>    all the three programs.
> 


The seed points are ok for white matter.
But, you are right, the current result looks different
from the one in the guide.  The changes in the anisotropic
diffusion filter make necessary to recalibrate the othere
parameters, probably the most critical are the Alpha and
Beta parameters of the sigmoid, since the front is leaking
close to the seed point, this lead to think that the
sigmod mapping is not really setting zero values boundaries
in the speed image. We will look at this.





> 3: Furthermore, I played with FastMarchingLevelSet.exe in
>    ITK\ITK1.2bin\InsightApplications-1.2.0\FastMarchingLevelSet\Debug. I
>    found that I could not see anything with VTK vender? Can you point
>    to me how to make it work? The image I tried is
> 
>        brainweb1e1a10f20.mha
>        brainweb1e1a10f20.raw
> 
 >


The VTK renderer is not functional at this point.
This is an open end (in the todo list). Since this
is pure visualization, it always goes to the bottom
of the priority list...



>
> 4: I also would like to play with the example of Deformable Models on
>    page 267, DeformableModel1.exe located in
>    ITK/ITK1.2bin/InsightToolkit-1.2.0/bin/Debug as given on page 272
> 
> ./ConfidenceConnected.exe ../../../../ITK1.2/InsightToolkit-1.2.0/Examples/Data/BrainProtonDensitySlice.png
>    ConfidenceConnectedOutPut1.png 81 114
> 
> I generated ConfidenceConnectedOutPut1.png with
> 
> ./ConfidenceConnected.exe ../../../../ITK1.2/InsightToolkit-1.2.0/Examples/Data/BrainProtonDensitySlice.png ConfidenceConnectedOutPut1.png 81 114
> 


Hmmmm, I don't understand your question here.
You are running the ConfidenceConnected executable which corresponds to
a region growing method. This is totally unrelated with the 
DeformableModel example....




> But I got an error (DeformableModel1Error1.bmp). If I press "retry",
> the same error comes. If I press "abort", I get another error
> (DeformableModel1Error2.bmp). Would you please let me know what went wrong?
> 


Are you running this from inside VisualStudio ?
Those are command line programs, you will have to run them from
a MS-DOS console, or a Cygwing shell, or....from VisualStudio if you
make sure to provide the parameters in the Project-Settings for
the Debug tab.

Are you passing any arguments to this program ?

It expects an input image and a BINARY image !
the binary image is a pre-segmentation of the input image.

This is an hybrid method in which a first segmentation
is used as prior for a second segmentation method.

It seems in any case that the text in this example
is pretty confusing and need some additional work.
We will look at it.


Thanks a lot for pointing this thing out.



Regards,


    Luis