MITK

Hi Ivo,
I've just downloaded from SVD the HEAD of mitk. I've seen you've changed
much, but I cannot find the UseMITK.cmake.in nor the UseMITK.cmake that =
is
supposed to be generated after the configuration step.
Result is that I cannot FIND_PACKAGE(MITK) now. It is supposed to work =
like
that, is the file is missing or is just supposed to work in any other =
way?

Best regards



 	Juan Antonio Moya
Computational Imaging Lab, Department of Technology - D. 313=20
Pompeu Fabra University, Pg. Circumval=B7laci=F3 8=20
08003 Barcelona - Spain=20
Phone +34 935 421348 =09

-----Mensaje original-----
De: mitk-users-bounces@...
[mailto:mitk-users-bounces@...] En nombre de Ivo Wolf
Enviado el: viernes, 25 de mayo de 2007 17:23
Para: mitk-users@...
Asunto: [mitk-users] Do NOT use FIND_PACKAGE(MITK) together
withFIND_PACKAGE(ITK)!

Dear all,

this is warning:

Do not use FIND_PACKAGE(MITK) together with FIND_PACKAGE(ITK)!

FIND_PACKAGE(MITK) will include ITK and VTK automatically,
so it is not necessary to include them separately.

Now the WARNING:
If you nevertheless do include ITK separately, you may experience with
removing nodes from the data-tree.

We will fix that problem as soon as possible.

Best regards,

Ivo


-------------------------------------------------------------------------=

This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
mitk-users mailing list
mitk-users@...
https://lists.sourceforge.net/lists/listinfo/mitk-users

Hi Juan,

you can find a MITKConfig.cmake file in your binary tree. This is 
supposed to be found and included by CMake's FIND_PACKAGE command. It 
sets a few variables like MITK_INCLUDE_DIRS and MITK_LINK_DIRECTORIES 
but you still have to write something like

INCLUDE_DIRECTORIES(${MITK_INCLUDE_DIRS})
LINK_DIRECTORIES(${MITK_LINK_DIRECTORIES})
TARGET_LINK_LIBRARIES(YourApp ${MITK_LIBRARIES})

in your own CMakeLists.txt . Maybe we will provide an UseMITK.cmake file 
doing this at some point. As Ivo already pointed out, there are still a 
few bugs with the MITKConfig.cmake so if you still have problems we 
would be happy to get some reports so we can fix them.

Regards,

Marco

Juan A. Moya Vicén schrieb:
> Hi Ivo,
> I've just downloaded from SVD the HEAD of mitk. I've seen you've changed
> much, but I cannot find the UseMITK.cmake.in nor the UseMITK.cmake that is
> supposed to be generated after the configuration step.
> Result is that I cannot FIND_PACKAGE(MITK) now. It is supposed to work like
> that, is the file is missing or is just supposed to work in any other way?
> 
> Best regards
> 
> 
> 
>  	Juan Antonio Moya
> Computational Imaging Lab, Department of Technology - D. 313 
> Pompeu Fabra University, Pg. Circumval·lació 8 
> 08003 Barcelona - Spain 
> Phone +34 935 421348 	
> 
> -----Mensaje original-----
> De: mitk-users-bounces@...
> [mailto:mitk-users-bounces@...] En nombre de Ivo Wolf
> Enviado el: viernes, 25 de mayo de 2007 17:23
> Para: mitk-users@...
> Asunto: [mitk-users] Do NOT use FIND_PACKAGE(MITK) together
> withFIND_PACKAGE(ITK)!
> 
> Dear all,
> 
> this is warning:
> 
> Do not use FIND_PACKAGE(MITK) together with FIND_PACKAGE(ITK)!
> 
> FIND_PACKAGE(MITK) will include ITK and VTK automatically,
> so it is not necessary to include them separately.
> 
> Now the WARNING:
> If you nevertheless do include ITK separately, you may experience with
> removing nodes from the data-tree.
> 
> We will fix that problem as soon as possible.
> 
> Best regards,
> 
> Ivo
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> mitk-users mailing list
> mitk-users@...
> https://lists.sourceforge.net/lists/listinfo/mitk-users
> 
> 
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> mitk-users mailing list
> mitk-users@...
> https://lists.sourceforge.net/lists/listinfo/mitk-users


-- 
----------------------------------------------------------------------
Dipl.-Inform. Med. Marco Nolden
Deutsches Krebsforschungszentrum       (German Cancer Research Center)
Div. Medical & Biological Informatics          Tel: (+49) 6221-42 2325
Im Neuenheimer Feld 280                        Fax: (+49) 6221-42 2345
D-69120 Heidelberg                             eMail: M.Nolden@...

Ok, I think I figured it out.  Now I am using the following function instead
of SetVolume:
  image->SetImportVolume( MyITKImage->GetBufferPointer(),0,0, mitk::Image::
ReferenceMemory );

It appears to call Modified on subsequent calls to the function, and updates
the internal memory.  Is this the best way to do this?

Thanks!
Jacob


On 5/25/07, Jacob Foshee <jacobf@...> wrote:
>
> I have an ITK Image which I am updating frequently.  I am using code like
> the following to display the image:
>                 //Initialize mitk image from itk
>                 mitk::Image::Pointer image = mitk::Image::New();
>                 image->InitializeByItk<ImageType>( GetImage() );
>                 image->SetVolume( GetImage()->GetBufferPointer() );
>
>                 //add the mitk image to the node
>                 mitk::DataTreeNode::Pointer node =
> mitk::DataTreeNode::New();
>                 node->SetData( image );
>                 mitk::DataTreePreOrderIterator it(m_Tree);
>                 it.Add(node);
>
> Obviously this is creating new images every time I update.  How can I just
> update the existing node w/out creating new copies of the image every time?
> Is there some way to do something like ITK: " myDataTreeNode.SetInput(
> MyImage->GetOutput() );" so it will update automatically?
>
> Thanks!
> Jacob
>
>

Hi again,
just to add:
There also is a mitkContourSetToPointSetFilter class!=20
Regards,
Ingmar
=20
________________________________

Von: Jacob Foshee [mailto:jacobf@...
Gesendet: Montag, 14. Mai 2007 22:02
An: Wegner Ingmar
Betreff: Re: [mitk-users] ContourInteractor
=20
Hi, Ingmar,
Sorry for my slow reply.  I was out of town and distracted by other =
projects.

Currently, I have adapted the ContourInteractor to create free-hand =
contours which are added to a ContourSet.  What I would like to do next =
is convert each contour to a filled raster representation, like a bit =
mask. =20

I have looked at vtk and itk, and believe that vtkPolyDataToImageStencil =
is the best match for the operation I would like to perform.  If that =
sounds reasonable to you, then my next question will be: what is the =
quickest way to convert the contour vertices to vtkPolyData?  or perhaps =
ContourSet to vtkPolyDataSet?=20

My next question regards rendering the contours as they are drawn.  I am =
getting a warning message that vtk is missing the Mapper to render the =
contour set in the 3D view.  If this is incomplete, I could work on that =
and share it, but I was having trouble finding where the geometry is =
defined for ContourSet.=20

Thanks,
Jacob
On 5/2/07, Ingmar Wegner <i.wegner@...> wrote:
Dear Jacob,
I have asked a few developer here at the division and a further class
you might be intrested in is the seedsinteractor.
(Sorry for the rare documentation but this class has been created by a
student who did his diploma thesis.)=20
This class provides drawing directly into a copy of an image. This
information (highlighted pixel) then can be superimposed on the image.
This was you should not have a problem on matching a position from a
pointset to the coordinates of the image, since the marked pixels have=20
the same coordinates than the original image.
Maybe this can help as well.

May I ask what you want to realize?
Maybe this makes things clearer and I can help to suggest what is
necessary to do in MITK.=20

Regards,
Ingmar

-------

Dipl.-Inform. Med. Ingmar Wegner    Tel.: (+49) 6221/42-3543
German Cancer Research Center       Fax.: (+49) 6221/42-2345
Div. Medical and Biological Informatics E130=20
Im Neuenheimer Feld 280               mailto:I.Wegner@...
D-69120 Heidelberg, Germany     http://www.dkfz.de/de/mbi/index.html
Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
-------




Jacob Foshee wrote:
> No, I didn't see the Extruded one. I will have a look at that as well. =

> Thanks.
>
> Here are the use cases I have currently worked up for our Contour tool
> development:
>
> Tool: Free Hand Creation
>
> Available on any 2D Cross-section view
>
> =9F Primary Mouse Down
> Start closed contour
>
> =9F Move with Primary Mouse Down
> Draw closed contour with starting point where mouse was first clicked
> and ending point at current mouse position. A straight line always=20
> connects starting and ending point as the mouse is moved showing that
> the contour is always closed.
>
> =9F Primary Mouse Release
> End closed contour.
> Rasterize closed contour and * union* with existing contours of same=20
> Contour Set.
> Update display on all views
>
> =9F Undo
> Removes last created contour and returns Contour Set to previous state
>
> Tool: Poly-line Creation
>
> Available on any 2D Cross-section view=20
>
> =9F Primary Mouse Release
> If Starting Contour, add first point to contour and Continue Contour
> If Continuing Contour, add next point to contour
>
> =9F Mouse Move
> If Continuing Contour, connect lines between cursor position and first =

> point and last point. Continue Contour.
>
> =9F Primary Double-Click
> Add point to contour
> Rasterize contour and * union* with existing contours of same Contour =
Set
> Update display on all views=20
> Return to Starting Contour mode.
>
> =9F Undo
> Removes last added point
> If last added point was final point, resumes editing of contour and
> returns Contour Set to previous state
>
> Tool: Nudge 2D Modification
>
> Available on any 2D Cross-section view
>
> =9F Behavior same as IKOEngelo
>
> =9F Primary Mouse Down
> If cursor position is over contour interior ( i.e. pixel is marked as
> covered by selected contour) then mode is Paint Contour
> Otherwise (cursor is over contour exterior) then mode is Erase Contour
>
> =9F Move with Primary Mouse Down
> Paint Contour Mode: Mark all pixels w/in Radius from cursor position=20
> as covered by selected contour
> Erase Contour Mode: Mark all pixels w/in Radius from cursor position
> as not covered by selected contour
>
> =9F Primary Mouse Release
> (If applicable) Update boundary representation of Contour=20
> Update all views
>
> Tool: Nudge 3D Modification
>
> =9F Works like Nudge 2D, except modifies multiple adjacent slices =
using
> a sphere or cylinder
>
> The first tool is very similar to what you have started with, so I=20
> would not mind getting that up to date. The Nudge tools require
> rasterizing the contour. But rendering and future
> boundary-manipulation tools will require a boundary rep (like you have
> now). So we will need to maintain both formats, or be able to quickly=20
> convert between the two.
> If you could direct me to existing code that helps with the
> rasterization/boundary-rep problem, we may be inclined to share our
> nudge tools.
> Thanks!
> Jacob=20
>
>
>
> On 4/26/07, *Ingmar Wegner* < i.wegner@...
> <mailto:i.wegner@... >> wrote:
>
>     Hi Jacob,
>     no, as far as I know, the mitkContourInteractor is not under
>     development
>     right now.
>     The last changes were done 5 months ago.=20
>     Right now we don't have any objectives to the ContourInteractor,
>     so any
>     changes are welcome. What requirements do you have?
>     There is another class, the ExtrudedContourCutter which is used in =
an=20
>     internal functionality. Did you check/recognize that one as well?
>
>     Thanks for the bugreport as well,
>     Ingmar
>
>     -------
>     Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543=20
>     German Cancer Research Center Fax.: (+49) 6221/42-2345
>     Div. Medical and Biological Informatics E130
>     Im Neuenheimer Feld 280 mailto:I.Wegner@...
>     <mailto:I.Wegner@...>
>     D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
>     < http://www.dkfz.de/de/mbi/index.html =
<http://www.dkfz.de/de/mbi/index.html> >
>     Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
>     < http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html =
<http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html> >
>     -------
>
>
>
>
>     Jacob Foshee wrote:
>     > I see. Thanks.
>     >
>     > I may be able to make some contributions to this class to get it =
in=20
>     > working order again. Is this part of MITK not currently under
>     active
>     > development? Might we discuss your objectives for a
>     ContourInteractor
>     > to see if they are similar to mine?=20
>     >
>     > On 4/25/07, *Ingmar Wegner* < i.wegner@...
>     <mailto:i.wegner@... >
>     > <mailto: i.wegner@...
>     <mailto:i.wegner@...>>> wrote:=20
>     >
>     > Hi Jacob,
>     > the Warnings are written because you switched on INTERDEBUG. The
>     > Warnings are sent by an itkWarningMarcro which writes the
>     > "Warning:" in=20
>     > front of the message. Every event you do which causes an action
>     also
>     > should bring up such a message. But this is just for =
debugging...
>     >
>     > You were right by switching to a different statemachine than the =

>     > pointsetinteractor-one. And yes, if the method
>     CalculateJurisdiction
>     > (sorry for that name) returns, that the statemachine can not =
cope
>     > with
>     > that event, then nothing happens. So overwriting it is correct.=20
>     >
>     > The class mitkContourInteractor has not been used for a while,
>     so any
>     > changes/fixes from you are very appreciated!
>     >
>     > If you have further problems or questions, don't hesitate to =
ask...=20
>     > Ingmar
>     >
>     > -------
>     > Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543
>     > German Cancer Research Center Fax.: (+49) 6221/42-2345
>     > Div. Medical and Biological Informatics E130=20
>     > Im Neuenheimer Feld 280 mailto: I.Wegner@...
>     <mailto:I.Wegner@...>
>     > <mailto: I.Wegner@... <mailto:I.Wegner@...>>
>     > D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
>     <http://www.dkfz.de/de/mbi/index.html>
>     > < http://www.dkfz.de/de/mbi/index.html>
>     > Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
>     > -------
>     >
>     >
>     >
>     >
>     > Jacob Foshee wrote:
>     > > I believe I found a couple of my problems. First, I switched
>     to use
>     > > the "PressMoveRelease" state machine. Second, I had to =
override=20
>     > the
>     > > default CalculateJurisdiction to allow for handling mouse
>     movement.
>     > >
>     > > Any additional advice would be appreciated.
>     > > Thanks!=20
>     > > Jacob
>     > >
>     > > On 4/24/07, *Jacob Foshee* < jacobf@...
>     <mailto:jacobf@... >
>     > <mailto: jacobf@... <mailto:jacobf@...>>
>     > > <mailto: jacobf@... <mailto:jacobf@...>  =
<mailto:jacobf@...> <mailto:
>     jacobf@... <mailto:jacobf@... >>>> wrote:
>     > >
>     > > Hello, all,
>     > > I am new to MITK, and have worked through the tutorials. I
>     > am now
>     > > trying to modify the 5th tutorial to switch from=20
>     > > PointSetInteractor to a ContourInteractor. However I can't =
seem
>     > > to add any new points to the contour.
>     > >
>     > > Defining INTERDEBUG I see:=20
>     > >
>     > > WARNING: In
>     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     > line 155
>     > > ContourInteractor (03D5BCC8): : Changing from State no=20
>     > points
>     > > or loaded points to State checkedN via Transition
>     > addPoint but
>     > > check n due to Event 3
>     > >
>     > >
>     > > I am not exactly sure what this means, or if it is the source =
of=20
>     > > the problem. Other warnings follow:
>     > >
>     > > WARNING: In
>     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     > line 188
>     > > ContourInteractor (03D5BCC8): Warning: no action defind for=20
>     > > Action (03BEDD10)
>     > > RTTI typeinfo: class mitk::Action
>     > > Reference Count: 2
>     > > Modified Time: 3158
>     > > Debug: Off
>     > > Observers:=20
>     > > none
>     > > in pointsetinteractor
>     > >
>     > > WARNING: In
>     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     > line 188=20
>     > > ContourInteractor (03D5BCC8): Warning: no action defind for
>     > > Action (03BEDD88)
>     > > RTTI typeinfo: class mitk::Action
>     > > Reference Count: 2
>     > > Modified Time: 3159
>     > > Debug: Off
>     > > Observers:
>     > > none
>     > > in pointsetinteractor
>     > >
>     > >
>     > > Thanks in advance. And also thanks for publicly sharing this
>     > > great library. It seems to be very well though-out and glues
>     > > together some good ideas.
>     > >=20
>     > > Cheers,
>     > > Jacob
>     > >
>     > >
>     > >
>     >
>     =
------------------------------------------------------------------------ =

>     >
>     > >
>     > >
>     >
>     =
-------------------------------------------------------------------------=

>     > > This SF.net email is sponsored by DB2 Express=20
>     > > Download DB2 Express C - the FREE version of DB2 express and =
take
>     > > control of your XML. No limits. Just data. Click to get it =
now.
>     > > http://sourceforge.net/powerbar/db2/
>     <http://sourceforge.net/powerbar/db2/>
>     > >
>     >
>     =
------------------------------------------------------------------------ =

>
>     >
>     > >
>     > > _______________________________________________
>     > > mitk-users mailing list
>     > > mitk-users@...
>     <mailto:mitk-users@...>
>     > <mailto: mitk-users@...
>     <mailto:mitk-users@...>>
>     > > https://lists.sourceforge.net/lists/listinfo/mitk-users
>     <https://lists.sourceforge.net/lists/listinfo/mitk-users>
>     > >=20
>     >
>     >
>
>
=20

I currently have an interactor which I refer to as FreeHandContourCreator.
It uses a ContourSet as its data node.  In the "Push" event, a new contour
is created and added to the ContourSet.

My goal now is to take contours in the ContourSet and create filled raster
representations.  That is, scan-convert to filled polygons.  The purpose is
that some contour operations, such as "painting" contours, are faster with
raster (pixel by pixel) representations than with boundary representations.
I am looking at both ITK and VTK for easy ways to scan-convert the contours=
,
but I am new to both.

My other problem is that the ContourSet is not being rendered.  I get this
error:
  vtkOpenGLPolyDataMapper (0242BA48): Mapper has no input.
I would like to try to use the mitkContourSetVTKMapper3D, as you suggested,
but I am not sure where to specify the mapper.  Could you help me there?

Thanks to everyone,
Jacob

On 5/15/07, Wegner Ingmar <I.Wegner@...> wrote:
>
>  Hi Jacob,
>
>
>
> I'm not completely sure, what you are going to implement!
>
> What shall the interactor realize? Drawing lines on a medical data that
> will be stencilled by the lines later on in order to eliminate artefacts =
or
> something?
>
> Please always send your mails to our sourceforge-list in order to help al=
l
> mitk-users and in order to archive the mails.
>
>
>
> Which data class do you use together with the ContourInteractor? An
> mitkPointSet?
>
> Then the PointSet should be easily convertible to a vtkPolyData since it
> is derived from a vtkPointSet. Converting the itkPointSet, which is
> wrapped by mitkPointSet, to vtkPointSet should also be no problem. Check
> ITK and VTK for this.
>
>
>
> For the renderers: It is dependent on the data, which is used by the
> interactor. The type of the data has to have a mitkMapper in 2D and 3D. A
> mitkPointSet has a mitkPointSetMapper2D and a mitkPointSetVtkMapper3D,
> which build up vtkPoints and connect them to the vtkRenderingPipeline. So
> which data do you use?
>
> For the mitkContourSet there are the two mappers mitkContourSetMapper2D
> and mitkContourSetVTKMapper3D.
>
>
>
>
>
> Regards ;),
>
> Ingmar
>
>
>
>
>  ------------------------------
>
> *Von:* Jacob Foshee [mailto:jacobf@...]
> *Gesendet:* Montag, 14. Mai 2007 22:02
> *An:* Wegner Ingmar
> *Betreff:* Re: [mitk-users] ContourInteractor
>
>
>
> Hi, Ingmar,
> Sorry for my slow reply.  I was out of town and distracted by other
> projects.
>
> Currently, I have adapted the ContourInteractor to create free-hand
> contours which are added to a ContourSet.  What I would like to do next i=
s
> convert each contour to a filled raster representation, like a bit mask.
>
> I have looked at vtk and itk, and believe that *vtkPolyDataToImageStencil=
*is the best match for the operation I would like to perform.  If that soun=
ds
> reasonable to you, then my next question will be: what is the quickest wa=
y
> to convert the contour vertices to vtkPolyData?  or perhaps ContourSet to
> vtkPolyDataSet?
>
> My next question regards rendering the contours as they are drawn.  I am
> getting a warning message that vtk is missing the Mapper to render the
> contour set in the 3D view.  If this is incomplete, I could work on that =
and
> share it, but I was having trouble finding where the geometry is defined =
for
> ContourSet.
>
> Thanks,
> Jacob
>
> On 5/2/07, *Ingmar Wegner* <i.wegner@...> wrote:
>
> Dear Jacob,
> I have asked a few developer here at the division and a further class
> you might be intrested in is the seedsinteractor.
> (Sorry for the rare documentation but this class has been created by a
> student who did his diploma thesis.)
> This class provides drawing directly into a copy of an image. This
> information (highlighted pixel) then can be superimposed on the image.
> This was you should not have a problem on matching a position from a
> pointset to the coordinates of the image, since the marked pixels have
> the same coordinates than the original image.
> Maybe this can help as well.
>
> May I ask what you want to realize?
> Maybe this makes things clearer and I can help to suggest what is
> necessary to do in MITK.
>
> Regards,
> Ingmar
>
> -------
>
> Dipl.-Inform. Med. Ingmar Wegner    Tel.: (+49) 6221/42-3543
> German Cancer Research Center       Fax.: (+49) 6221/42-2345
> Div. Medical and Biological Informatics E130
> Im Neuenheimer Feld 280               mailto:I.Wegner@...
> D-69120 Heidelberg, Germany     http://www.dkfz.de/de/mbi/index.html
> Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
> -------
>
>
>
>
> Jacob Foshee wrote:
> > No, I didn't see the Extruded one. I will have a look at that as well.
> > Thanks.
> >
> > Here are the use cases I have currently worked up for our Contour tool
> > development:
> >
> > Tool: Free Hand Creation
> >
> > Available on any 2D Cross-section view
> >
> > =9F Primary Mouse Down
> > Start closed contour
> >
> > =9F Move with Primary Mouse Down
> > Draw closed contour with starting point where mouse was first clicked
> > and ending point at current mouse position. A straight line always
> > connects starting and ending point as the mouse is moved showing that
> > the contour is always closed.
> >
> > =9F Primary Mouse Release
> > End closed contour.
> > Rasterize closed contour and * union* with existing contours of same
> > Contour Set.
> > Update display on all views
> >
> > =9F Undo
> > Removes last created contour and returns Contour Set to previous state
> >
> > Tool: Poly-line Creation
> >
> > Available on any 2D Cross-section view
> >
> > =9F Primary Mouse Release
> > If Starting Contour, add first point to contour and Continue Contour
> > If Continuing Contour, add next point to contour
> >
> > =9F Mouse Move
> > If Continuing Contour, connect lines between cursor position and first
> > point and last point. Continue Contour.
> >
> > =9F Primary Double-Click
> > Add point to contour
> > Rasterize contour and * union* with existing contours of same Contour
> Set
> > Update display on all views
> > Return to Starting Contour mode.
> >
> > =9F Undo
> > Removes last added point
> > If last added point was final point, resumes editing of contour and
> > returns Contour Set to previous state
> >
> > Tool: Nudge 2D Modification
> >
> > Available on any 2D Cross-section view
> >
> > =9F Behavior same as IKOEngelo
> >
> > =9F Primary Mouse Down
> > If cursor position is over contour interior ( i.e. pixel is marked as
> > covered by selected contour) then mode is Paint Contour
> > Otherwise (cursor is over contour exterior) then mode is Erase Contour
> >
> > =9F Move with Primary Mouse Down
> > Paint Contour Mode: Mark all pixels w/in Radius from cursor position
> > as covered by selected contour
> > Erase Contour Mode: Mark all pixels w/in Radius from cursor position
> > as not covered by selected contour
> >
> > =9F Primary Mouse Release
> > (If applicable) Update boundary representation of Contour
> > Update all views
> >
> > Tool: Nudge 3D Modification
> >
> > =9F Works like Nudge 2D, except modifies multiple adjacent slices using
> > a sphere or cylinder
> >
> > The first tool is very similar to what you have started with, so I
> > would not mind getting that up to date. The Nudge tools require
> > rasterizing the contour. But rendering and future
> > boundary-manipulation tools will require a boundary rep (like you have
> > now). So we will need to maintain both formats, or be able to quickly
> > convert between the two.
> > If you could direct me to existing code that helps with the
> > rasterization/boundary-rep problem, we may be inclined to share our
> > nudge tools.
> > Thanks!
> > Jacob
> >
> >
> >
> > On 4/26/07, *Ingmar Wegner* < i.wegner@...
> > <mailto:i.wegner@... >> wrote:
> >
> >     Hi Jacob,
> >     no, as far as I know, the mitkContourInteractor is not under
> >     development
> >     right now.
> >     The last changes were done 5 months ago.
> >     Right now we don't have any objectives to the ContourInteractor,
> >     so any
> >     changes are welcome. What requirements do you have?
> >     There is another class, the ExtrudedContourCutter which is used in
> an
> >     internal functionality. Did you check/recognize that one as well?
> >
> >     Thanks for the bugreport as well,
> >     Ingmar
> >
> >     -------
> >     Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543
> >     German Cancer Research Center Fax.: (+49) 6221/42-2345
> >     Div. Medical and Biological Informatics E130
> >     Im Neuenheimer Feld 280 mailto:I.Wegner@...
> >     <mailto:I.Wegner@...>
> >     D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
> >     < http://www.dkfz.de/de/mbi/index.html>
> >     Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
> >     < http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html>
> >     -------
> >
> >
> >
> >
> >     Jacob Foshee wrote:
> >     > I see. Thanks.
> >     >
> >     > I may be able to make some contributions to this class to get it
> in
> >     > working order again. Is this part of MITK not currently under
> >     active
> >     > development? Might we discuss your objectives for a
> >     ContourInteractor
> >     > to see if they are similar to mine?
> >     >
> >     > On 4/25/07, *Ingmar Wegner* < i.wegner@...
> >     <mailto:i.wegner@... >
> >     > <mailto: i.wegner@...
> >     <mailto:i.wegner@...>>> wrote:
> >     >
> >     > Hi Jacob,
> >     > the Warnings are written because you switched on INTERDEBUG. The
> >     > Warnings are sent by an itkWarningMarcro which writes the
> >     > "Warning:" in
> >     > front of the message. Every event you do which causes an action
> >     also
> >     > should bring up such a message. But this is just for debugging...
> >     >
> >     > You were right by switching to a different statemachine than the
> >     > pointsetinteractor-one. And yes, if the method
> >     CalculateJurisdiction
> >     > (sorry for that name) returns, that the statemachine can not cope
> >     > with
> >     > that event, then nothing happens. So overwriting it is correct.
> >     >
> >     > The class mitkContourInteractor has not been used for a while,
> >     so any
> >     > changes/fixes from you are very appreciated!
> >     >
> >     > If you have further problems or questions, don't hesitate to
> ask...
> >     > Ingmar
> >     >
> >     > -------
> >     > Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543
> >     > German Cancer Research Center Fax.: (+49) 6221/42-2345
> >     > Div. Medical and Biological Informatics E130
> >     > Im Neuenheimer Feld 280 mailto: I.Wegner@...
> >     <mailto:I.Wegner@...>
> >     > <mailto: I.Wegner@... <mailto:I.Wegner@...>>
> >     > D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
> >     <http://www.dkfz.de/de/mbi/index.html>
> >     > < http://www.dkfz.de/de/mbi/index.html>
> >     > Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
> >     > -------
> >     >
> >     >
> >     >
> >     >
> >     > Jacob Foshee wrote:
> >     > > I believe I found a couple of my problems. First, I switched
> >     to use
> >     > > the "PressMoveRelease" state machine. Second, I had to override
> >     > the
> >     > > default CalculateJurisdiction to allow for handling mouse
> >     movement.
> >     > >
> >     > > Any additional advice would be appreciated.
> >     > > Thanks!
> >     > > Jacob
> >     > >
> >     > > On 4/24/07, *Jacob Foshee* < jacobf@...
> >     <mailto:jacobf@... >
> >     > <mailto: jacobf@... <mailto:jacobf@...>>
> >     > > <mailto: jacobf@... <mailto:jacobf@...> <mailto:
> >     jacobf@... <mailto:jacobf@... >>>> wrote:
> >     > >
> >     > > Hello, all,
> >     > > I am new to MITK, and have worked through the tutorials. I
> >     > am now
> >     > > trying to modify the 5th tutorial to switch from
> >     > > PointSetInteractor to a ContourInteractor. However I can't seem
> >     > > to add any new points to the contour.
> >     > >
> >     > > Defining INTERDEBUG I see:
> >     > >
> >     > > WARNING: In
> >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
> >     > line 155
> >     > > ContourInteractor (03D5BCC8): : Changing from State no
> >     > points
> >     > > or loaded points to State checkedN via Transition
> >     > addPoint but
> >     > > check n due to Event 3
> >     > >
> >     > >
> >     > > I am not exactly sure what this means, or if it is the source o=
f
>
> >     > > the problem. Other warnings follow:
> >     > >
> >     > > WARNING: In
> >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
> >     > line 188
> >     > > ContourInteractor (03D5BCC8): Warning: no action defind for
> >     > > Action (03BEDD10)
> >     > > RTTI typeinfo: class mitk::Action
> >     > > Reference Count: 2
> >     > > Modified Time: 3158
> >     > > Debug: Off
> >     > > Observers:
> >     > > none
> >     > > in pointsetinteractor
> >     > >
> >     > > WARNING: In
> >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
> >     > line 188
> >     > > ContourInteractor (03D5BCC8): Warning: no action defind for
> >     > > Action (03BEDD88)
> >     > > RTTI typeinfo: class mitk::Action
> >     > > Reference Count: 2
> >     > > Modified Time: 3159
> >     > > Debug: Off
> >     > > Observers:
> >     > > none
> >     > > in pointsetinteractor
> >     > >
> >     > >
> >     > > Thanks in advance. And also thanks for publicly sharing this
> >     > > great library. It seems to be very well though-out and glues
> >     > > together some good ideas.
> >     > >
> >     > > Cheers,
> >     > > Jacob
> >     > >
> >     > >
> >     > >
> >     >
> >
> ------------------------------------------------------------------------
> >     >
> >     > >
> >     > >
> >     >
> >
> -------------------------------------------------------------------------
> >     > > This SF.net email is sponsored by DB2 Express
> >     > > Download DB2 Express C - the FREE version of DB2 express and
> take
> >     > > control of your XML. No limits. Just data. Click to get it now.
> >     > > http://sourceforge.net/powerbar/db2/
> >     <http://sourceforge.net/powerbar/db2/>
> >     > >
> >     >
> >
> ------------------------------------------------------------------------
> >
> >     >
> >     > >
> >     > > _______________________________________________
> >     > > mitk-users mailing list
> >     > > mitk-users@...
> >     <mailto:mitk-users@...>
> >     > <mailto: mitk-users@...
> >     <mailto:mitk-users@...>>
> >     > > https://lists.sourceforge.net/lists/listinfo/mitk-users
> >     <https://lists.sourceforge.net/lists/listinfo/mitk-users>
> >     > >
> >     >
> >     >
> >
> >
>
>
>

Hi Jacob,
now I see! Thanks!
Usually the mapper should be added automatically by the 
mitkCoreObjectFactory (line 241 and 300). Here a mitkContourMapper2D and 
a mitkContourSetVtkMapper3D gets added to the dataTreeNode in the 
appropriate slot (slot 1 for 2D and slot 2 for 3D).

So the regular steps for creating a new DataTreeNode with an interactor, 
mapper and data are:
*create data (ContourSet)
    mitk::ContourSet::Pointer contourSet = mitk::PointSet::New();

*create DataTreeNode
    mitk::DataTreeNode myNewNode = mitk::DataTreeNode::New();

*set data to DataTreeNode
    myNewNode->SetData(contourSet);

*create an Interactor
    mitk::MySpecialContourInteractor::Pointer contourInteractor = 
mitk::MySpecialContourInteractor::New(...);

*and add it to the dataTreeNode. Here the data has to already exist.
    myNewNode->SetInteractor(contourInteractor);

*set desired properties and default properties
    myNewNode->GetPropertyList()->SetProperty("layer",new 
mitk::IntProperty(1));//properties like name
    myNewNode->SetColor(1.0,0.0,0.0);//and color.

*then you can force to use a special Mapper
    mitk::ContourSetVtkMapper3D::Pointer mapper3D = 
mitk::ContourSetVtkMapper3D::New();
    myNewNode->SetMapper( 2, mapper3D );

*and then add to the dataTreeNode
    mitk::DataTreeIteratorClone child = 
this->GetDataTreeIterator()->GetTree();
    child->GoToChild(child->CountChildren()-1);//go to the next child
    child->Add(myNewNode);


You shouldn't have to define any Vtk-classes. The error " 
vtkOpenGLPolyDataMapper (0242BA48): Mapper has no input." sounds like a 
not correctly initialized DataTreeNode. If this doesn't solve anything, 
then send an extract of your code.


Best regards,
Ingmar


-------
Dipl.-Inform. Med. Ingmar Wegner    Tel.: (+49) 6221/42-3543  
German Cancer Research Center       Fax.: (+49) 6221/42-2345  
Div. Medical and Biological Informatics E130                      
Im Neuenheimer Feld 280               mailto:I.Wegner@...   
D-69120 Heidelberg, Germany  http://www.dkfz.de/de/mbi/index.html
Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
-------




Jacob Foshee wrote:
> I currently have an interactor which I refer to as 
> FreeHandContourCreator.  It uses a ContourSet as its data node.  In 
> the "Push" event, a new contour is created and added to the ContourSet.
>
> My goal now is to take contours in the ContourSet and create filled 
> raster representations.  That is, scan-convert to filled polygons.  
> The purpose is that some contour operations, such as "painting" 
> contours, are faster with raster (pixel by pixel) representations than 
> with boundary representations.  I am looking at both ITK and VTK for 
> easy ways to scan-convert the contours, but I am new to both.
>
> My other problem is that the ContourSet is not being rendered.  I get 
> this error:
>   vtkOpenGLPolyDataMapper (0242BA48): Mapper has no input.
> I would like to try to use the mitkContourSetVTKMapper3D, as you 
> suggested, but I am not sure where to specify the mapper.  Could you 
> help me there?
>
> Thanks to everyone,
> Jacob
>
> On 5/15/07, *Wegner Ingmar* <I.Wegner@... 
> <mailto:I.Wegner@...>> wrote:
>
>     Hi Jacob,
>
>      
>
>     I'm not completely sure, what you are going to implement!
>
>     What shall the interactor realize? Drawing lines on a medical data
>     that will be stencilled by the lines later on in order to
>     eliminate artefacts or something?
>
>     Please always send your mails to our sourceforge-list in order to
>     help all mitk-users and in order to archive the mails.
>
>      
>
>     Which data class do you use together with the ContourInteractor?
>     An mitkPointSet?
>
>     Then the PointSet should be easily convertible to a vtkPolyData
>     since it is derived from a vtkPointSet. Converting the
>     itkPointSet, which is wrapped by mitkPointSet, to vtkPointSet
>     should also be no problem. Check ITK and VTK for this.
>
>      
>
>     For the renderers: It is dependent on the data, which is used by
>     the interactor. The type of the data has to have a mitkMapper in
>     2D and 3D. A mitkPointSet has a mitkPointSetMapper2D and a
>     mitkPointSetVtkMapper3D, which build up vtkPoints and connect them
>     to the vtkRenderingPipeline. So which data do you use?
>
>     For the mitkContourSet there are the two mappers
>     mitkContourSetMapper2D and mitkContourSetVTKMapper3D.
>
>      
>
>      
>
>     Regards ;),
>
>     Ingmar
>
>      
>
>      
>
>     ------------------------------------------------------------------------
>
>     *Von:* Jacob Foshee [mailto:jacobf@...
>     <mailto:jacobf@...>]
>     *Gesendet:* Montag, 14. Mai 2007 22:02
>     *An:* Wegner Ingmar
>     *Betreff:* Re: [mitk-users] ContourInteractor
>
>      
>
>     Hi, Ingmar,
>     Sorry for my slow reply.  I was out of town and distracted by
>     other projects.
>
>     Currently, I have adapted the ContourInteractor to create
>     free-hand contours which are added to a ContourSet.  What I would
>     like to do next is convert each contour to a filled raster
>     representation, like a bit mask. 
>
>     I have looked at vtk and itk, and believe that
>     *vtkPolyDataToImageStencil* is the best match for the operation I
>     would like to perform.  If that sounds reasonable to you, then my
>     next question will be: what is the quickest way to convert the
>     contour vertices to vtkPolyData?  or perhaps ContourSet to
>     vtkPolyDataSet?
>
>     My next question regards rendering the contours as they are
>     drawn.  I am getting a warning message that vtk is missing the
>     Mapper to render the contour set in the 3D view.  If this is
>     incomplete, I could work on that and share it, but I was having
>     trouble finding where the geometry is defined for ContourSet.
>
>     Thanks,
>     Jacob
>
>     On 5/2/07, *Ingmar Wegner* <i.wegner@...
>     <mailto:i.wegner@...>> wrote:
>
>     Dear Jacob,
>     I have asked a few developer here at the division and a further class
>     you might be intrested in is the seedsinteractor.
>     (Sorry for the rare documentation but this class has been created by a
>     student who did his diploma thesis.)
>     This class provides drawing directly into a copy of an image. This
>     information (highlighted pixel) then can be superimposed on the image.
>     This was you should not have a problem on matching a position from a
>     pointset to the coordinates of the image, since the marked pixels
>     have
>     the same coordinates than the original image.
>     Maybe this can help as well.
>
>     May I ask what you want to realize?
>     Maybe this makes things clearer and I can help to suggest what is
>     necessary to do in MITK.
>
>     Regards,
>     Ingmar
>
>     -------
>
>     Dipl.-Inform. Med. Ingmar Wegner    Tel.: (+49) 6221/42-3543
>     German Cancer Research Center       Fax.: (+49) 6221/42-2345
>     Div. Medical and Biological Informatics E130
>     Im Neuenheimer Feld 280               mailto:I.Wegner@...
>     <mailto:I.Wegner@...>
>     D-69120 Heidelberg, Germany     http://www.dkfz.de/de/mbi/index.html
>     <http://www.dkfz.de/de/mbi/index.html>
>     Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
>     -------
>
>
>
>
>     Jacob Foshee wrote:
>     > No, I didn't see the Extruded one. I will have a look at that as
>     well.
>     > Thanks.
>     >
>     > Here are the use cases I have currently worked up for our Contour
>     tool
>     > development:
>     >
>     > Tool: Free Hand Creation
>     >
>     > Available on any 2D Cross-section view
>     >
>     > Ÿ Primary Mouse Down
>     > Start closed contour
>     >
>     > Ÿ Move with Primary Mouse Down
>     > Draw closed contour with starting point where mouse was first clicked
>     > and ending point at current mouse position. A straight line always
>     > connects starting and ending point as the mouse is moved showing that
>     > the contour is always closed.
>     >
>     > Ÿ Primary Mouse Release
>     > End closed contour.
>     > Rasterize closed contour and * union* with existing contours of same
>     > Contour Set.
>     > Update display on all views
>     >
>     > Ÿ Undo
>     > Removes last created contour and returns Contour Set to previous
>     state
>     >
>     > Tool: Poly-line Creation
>     >
>     > Available on any 2D Cross-section view
>     >
>     > Ÿ Primary Mouse Release
>     > If Starting Contour, add first point to contour and Continue Contour
>     > If Continuing Contour, add next point to contour
>     >
>     > Ÿ Mouse Move
>     > If Continuing Contour, connect lines between cursor position and
>     first
>     > point and last point. Continue Contour.
>     >
>     > Ÿ Primary Double-Click
>     > Add point to contour
>     > Rasterize contour and * union* with existing contours of same
>     Contour Set
>     > Update display on all views
>     > Return to Starting Contour mode.
>     >
>     > Ÿ Undo
>     > Removes last added point
>     > If last added point was final point, resumes editing of contour and
>     > returns Contour Set to previous state
>     >
>     > Tool: Nudge 2D Modification
>     >
>     > Available on any 2D Cross-section view
>     >
>     > Ÿ Behavior same as IKOEngelo
>     >
>     > Ÿ Primary Mouse Down
>     > If cursor position is over contour interior ( i.e. pixel is marked as
>     > covered by selected contour) then mode is Paint Contour
>     > Otherwise (cursor is over contour exterior) then mode is Erase
>     Contour
>     >
>     > Ÿ Move with Primary Mouse Down
>     > Paint Contour Mode: Mark all pixels w/in Radius from cursor position
>     > as covered by selected contour
>     > Erase Contour Mode: Mark all pixels w/in Radius from cursor position
>     > as not covered by selected contour
>     >
>     > Ÿ Primary Mouse Release
>     > (If applicable) Update boundary representation of Contour
>     > Update all views
>     >
>     > Tool: Nudge 3D Modification
>     >
>     > Ÿ Works like Nudge 2D, except modifies multiple adjacent slices using
>     > a sphere or cylinder
>     >
>     > The first tool is very similar to what you have started with, so I
>     > would not mind getting that up to date. The Nudge tools require
>     > rasterizing the contour. But rendering and future
>     > boundary-manipulation tools will require a boundary rep (like you
>     have
>     > now). So we will need to maintain both formats, or be able to
>     quickly
>     > convert between the two.
>     > If you could direct me to existing code that helps with the
>     > rasterization/boundary-rep problem, we may be inclined to share our
>     > nudge tools.
>     > Thanks!
>     > Jacob
>     >
>     >
>     >
>     > On 4/26/07, *Ingmar Wegner* < i.wegner@...
>     <mailto:i.wegner@...>
>     > <mailto:i.wegner@...
>     <mailto:i.wegner@...>>> wrote:
>     >
>     >     Hi Jacob,
>     >     no, as far as I know, the mitkContourInteractor is not under
>     >     development
>     >     right now.
>     >     The last changes were done 5 months ago.
>     >     Right now we don't have any objectives to the ContourInteractor,
>     >     so any
>     >     changes are welcome. What requirements do you have?
>     >     There is another class, the ExtrudedContourCutter which is
>     used in an
>     >     internal functionality. Did you check/recognize that one as well?
>     >
>     >     Thanks for the bugreport as well,
>     >     Ingmar
>     >
>     >     -------
>     >     Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543
>     >     German Cancer Research Center Fax.: (+49) 6221/42-2345
>     >     Div. Medical and Biological Informatics E130
>     >     Im Neuenheimer Feld 280 mailto:I.Wegner@...
>     <mailto:I.Wegner@...>
>     >     <mailto:I.Wegner@... <mailto:I.Wegner@...>>
>     >     D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
>     <http://www.dkfz.de/de/mbi/index.html>
>     >     < http://www.dkfz.de/de/mbi/index.html>
>     >     Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
>     >     < http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html>
>     >     -------
>     >
>     >
>     >
>     >
>     >     Jacob Foshee wrote:
>     >     > I see. Thanks.
>     >     >
>     >     > I may be able to make some contributions to this class to
>     get it in
>     >     > working order again. Is this part of MITK not currently under
>     >     active
>     >     > development? Might we discuss your objectives for a
>     >     ContourInteractor
>     >     > to see if they are similar to mine?
>     >     >
>     >     > On 4/25/07, *Ingmar Wegner* < i.wegner@...
>     <mailto:i.wegner@...>
>     >     <mailto:i.wegner@...
>     <mailto:i.wegner@...>>
>     >     > <mailto: i.wegner@...
>     <mailto:i.wegner@...>
>     >     <mailto:i.wegner@...
>     <mailto:i.wegner@...>>>> wrote:
>     >     >
>     >     > Hi Jacob,
>     >     > the Warnings are written because you switched on
>     INTERDEBUG. The
>     >     > Warnings are sent by an itkWarningMarcro which writes the
>     >     > "Warning:" in
>     >     > front of the message. Every event you do which causes an action
>     >     also
>     >     > should bring up such a message. But this is just for
>     debugging...
>     >     >
>     >     > You were right by switching to a different statemachine
>     than the
>     >     > pointsetinteractor-one. And yes, if the method
>     >     CalculateJurisdiction
>     >     > (sorry for that name) returns, that the statemachine can
>     not cope
>     >     > with
>     >     > that event, then nothing happens. So overwriting it is
>     correct.
>     >     >
>     >     > The class mitkContourInteractor has not been used for a while,
>     >     so any
>     >     > changes/fixes from you are very appreciated!
>     >     >
>     >     > If you have further problems or questions, don't hesitate
>     to ask...
>     >     > Ingmar
>     >     >
>     >     > -------
>     >     > Dipl.-Inform. Med. Ingmar Wegner Tel.: (+49) 6221/42-3543
>     >     > German Cancer Research Center Fax.: (+49) 6221/42-2345
>     >     > Div. Medical and Biological Informatics E130
>     >     > Im Neuenheimer Feld 280 mailto: I.Wegner@...
>     <mailto:I.Wegner@...>
>     >     <mailto:I.Wegner@... <mailto:I.Wegner@...>>
>     >     > <mailto: I.Wegner@... <mailto:I.Wegner@...>
>     <mailto:I.Wegner@... <mailto:I.Wegner@...>>>
>     >     > D-69120 Heidelberg, Germany http://www.dkfz.de/de/mbi/index.html
>     <http://www.dkfz.de/de/mbi/index.html>
>     >     <http://www.dkfz.de/de/mbi/index.html>
>     >     > < http://www.dkfz.de/de/mbi/index.html>
>     >     > Homepage: http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html
>     >     > -------
>     >     >
>     >     >
>     >     >
>     >     >
>     >     > Jacob Foshee wrote:
>     >     > > I believe I found a couple of my problems. First, I switched
>     >     to use
>     >     > > the "PressMoveRelease" state machine. Second, I had to
>     override
>     >     > the
>     >     > > default CalculateJurisdiction to allow for handling mouse
>     >     movement.
>     >     > >
>     >     > > Any additional advice would be appreciated.
>     >     > > Thanks!
>     >     > > Jacob
>     >     > >
>     >     > > On 4/24/07, *Jacob Foshee* < jacobf@...
>     <mailto:jacobf@...>
>     >     <mailto:jacobf@... <mailto:jacobf@...>>
>     >     > <mailto: jacobf@... <mailto:jacobf@...>
>     <mailto:jacobf@... <mailto:jacobf@...>>>
>     >     > > <mailto: jacobf@... <mailto:jacobf@...>
>     <mailto: jacobf@... <mailto:jacobf@...>> <mailto:
>     >     jacobf@... <mailto:jacobf@...>
>     <mailto:jacobf@... <mailto:jacobf@...>>>>> wrote:
>     >     > >
>     >     > > Hello, all,
>     >     > > I am new to MITK, and have worked through the tutorials. I
>     >     > am now
>     >     > > trying to modify the 5th tutorial to switch from
>     >     > > PointSetInteractor to a ContourInteractor. However I
>     can't seem
>     >     > > to add any new points to the contour.
>     >     > >
>     >     > > Defining INTERDEBUG I see:
>     >     > >
>     >     > > WARNING: In
>     >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     >     > line 155
>     >     > > ContourInteractor (03D5BCC8): : Changing from State no
>     >     > points
>     >     > > or loaded points to State checkedN via Transition
>     >     > addPoint but
>     >     > > check n due to Event 3
>     >     > >
>     >     > >
>     >     > > I am not exactly sure what this means, or if it is the
>     source of
>     >     > > the problem. Other warnings follow:
>     >     > >
>     >     > > WARNING: In
>     >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     >     > line 188
>     >     > > ContourInteractor (03D5BCC8): Warning: no action defind for
>     >     > > Action (03BEDD10)
>     >     > > RTTI typeinfo: class mitk::Action
>     >     > > Reference Count: 2
>     >     > > Modified Time: 3158
>     >     > > Debug: Off
>     >     > > Observers:
>     >     > > none
>     >     > > in pointsetinteractor
>     >     > >
>     >     > > WARNING: In
>     >     > > 3rdParty\MITK\Core\Interactions\mitkStateMachine.cpp,
>     >     > line 188
>     >     > > ContourInteractor (03D5BCC8): Warning: no action defind for
>     >     > > Action (03BEDD88)
>     >     > > RTTI typeinfo: class mitk::Action
>     >     > > Reference Count: 2
>     >     > > Modified Time: 3159
>     >     > > Debug: Off
>     >     > > Observers:
>     >     > > none
>     >     > > in pointsetinteractor
>     >     > >
>     >     > >
>     >     > > Thanks in advance. And also thanks for publicly sharing this
>     >     > > great library. It seems to be very well though-out and glues
>     >     > > together some good ideas.
>     >     > >
>     >     > > Cheers,
>     >     > > Jacob
>     >     > >
>     >     > >
>     >     > >
>     >     >
>     >    
>     ------------------------------------------------------------------------
>
>     >     >
>     >     > >
>     >     > >
>     >     >
>     >    
>     -------------------------------------------------------------------------
>     >     > > This SF.net email is sponsored by DB2 Express
>     >     > > Download DB2 Express C - the FREE version of DB2 express
>     and take
>     >     > > control of your XML. No limits. Just data. Click to get
>     it now.
>     >     > > http://sourceforge.net/powerbar/db2/
>     >     <http://sourceforge.net/powerbar/db2/>
>     >     > >
>     >     >
>     >    
>     ------------------------------------------------------------------------
>
>     >
>     >     >
>     >     > >
>     >     > > _______________________________________________
>     >     > > mitk-users mailing list
>     >     > > mitk-users@...
>     <mailto:mitk-users@...>
>     >     <mailto:mitk-users@...
>     <mailto:mitk-users@...>>
>     >     > <mailto: mitk-users@...
>     <mailto:mitk-users@...>
>     >     <mailto:mitk-users@...
>     <mailto:mitk-users@...>>>
>     >     > > https://lists.sourceforge.net/lists/listinfo/mitk-users
>     >     <https://lists.sourceforge.net/lists/listinfo/mitk-users>
>     >     > >
>     >     >
>     >     >
>     >
>     >
>
>      
>
>

<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
  <meta content="text/html;charset=ISO-8859-1" http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
Hi Petar,<br>
you are right, the according stereo-rendering in VTK has to switched
on, <br>
but the QT-Widget has to be prepared as well.<br>
See the class QmitkSelectableGLEidget.ui.h line 54 and the
documentation above together with QT documentation of the <a
 href="http://doc.trolltech.com/3.3/qglwidget.html">QGLWidget</a> and <a
 href="http://doc.trolltech.com/3.3/qglformat.html">QGLFormat</a>.<br>
Check to also pass QGL::Stereo to the constructor of QmitkRenderWindow
since stereo-rendering is disabled by default in QT.<br>
<br>
By the way, we appreciate any code from you that adds features to MITK.<br>
<br>
Best regards,<br>
Ingmar<br>
<br>
<pre class="moz-signature" cols="72">-------
Dipl.-Inform. Med. Ingmar Wegner    Tel.: (+49) 6221/42-3543  
German Cancer Research Center       Fax.: (+49) 6221/42-2345  
Div. Medical and Biological Informatics E130                      
Im Neuenheimer Feld 280               <a class="moz-txt-link-freetext" href="mailto:I.Wegner@...>   
D-69120 Heidelberg, Germany 	<a class="moz-txt-link-abbreviated" href="http://www.dkfz.de/de/mbi/index.html">www.dkfz.de/de/mbi/index.html</a>
Homepage: <a class="moz-txt-link-freetext" href="http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html">http://www.dkfz.de/de/mbi/people/Ingmar_Wegner.html</a>
-------

</pre>
<br>
<br>
kompjuterska radionica wrote:
<blockquote
 cite="mid2234d5070705151901p6e05d60jfc6cf832eb72d577@..."
 type="cite">Hi everyone,<br>
  <br>
I'm trying to add active stereo to MITKSampleApp, but having some
problems.<br>
System is suse 10.2 with Quadra FX 4500, and I can see active stereo in
  <br>
another application. <br>
  <br>
I've changed a line in Applications/MITKSampleApp/SampleApp.cpp:
  <br>
&nbsp; vtkrenderwindow-&gt;SetStereoTypeToRedBlue();<br>
to<br>
&nbsp; vtkrenderwindow-&gt;SetStereoTypeToCrystalEyes();<br>
  <br>
which gave me the following error when switching to stereo in the
application:<br>
&nbsp; VtkRenderWindow (0x96cf970): Adjusting stereo mode on a window that <br>
&nbsp; does not support stereo type CrystalEyes is not possible<br>
&nbsp; <br>
  <br>
Then I added the following to
AppModules/QmitkStdMultiWidget/QmitkStdMultiWidget.ui.h (at the&nbsp; init):<br>
&nbsp; vtkRenderWindow * vtkrenderwindow =<br>
&nbsp; (vtkRenderWindow*)((mitk::OpenGLRenderer*)
  <br>
&nbsp;
(m_MultiWidget-&gt;mitkWidget4-&gt;GetRenderer()))-&gt;GetVtkRenderWindow();<br>
&nbsp; vtkrenderwindow-&gt;SetStereoCapableWindow(1);<br>
  <br>
which gave me no error when I switch to stereo, but the image in&nbsp; 3d
widget shifts a few pixels to the
  <br>
left and that's all -&nbsp; no stereo.<br>
  <br>
  <br>
Best regards,<br>
Petar<br>
  <br>
  <pre wrap="">
<hr size="4" width="90%">
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
<a class="moz-txt-link-freetext" href="http://sourceforge.net/powerbar/db2/">http://sourceforge.net/powerbar/db2/</a></pre>
  <pre wrap="">
<hr size="4" width="90%">
_______________________________________________
mitk-users mailing list
<a class="moz-txt-link-abbreviated" href="mailto:mitk-users@...>
<a class="moz-txt-link-freetext" href="https://lists.sourceforge.net/lists/listinfo/mitk-users">https://lists.sourceforge.net/lists/listinfo/mitk-users</a>
  </pre>
</blockquote>
</body>
</html>

Hello,

Besides your cmake error, you'll have another problem:

> DESIRED_QT_VERSION               4

"Qt 4.x will not be supported in the near future."  
(http://mbi.dkfz-heidelberg.de/mitk/download/download.html)
(Last modified: 2006-07-04)

When will Qt 4.x support be available? Will it ever be available? Is "not  
in the near future" still valid?

Thanks,
Wolfgang




On Thu, 03 May 2007 17:18:34 +0200, Mark Alexiuk <malexiuk@...>  
wrote:

> Hi,
>
> I would appreciate any help on the following cmake error during a MITK  
> build
> attempt.
>
> The build MITK (0.6) uses cygwin (and gcc 3.4.4) on Windows XP.
> Since it's an out-of-source build, I run ccmake from the build directory.
> I've turned off all the cmake settings I can think of (see listing  
> below).
>
> Thanks in advance.
> Trying to enjoy ITK,
> Mark
>
> -------------
> Here is my error:
>
>  CMake Error: Error in cmake code at
>  /cygdrive/c/MITK-cvs/mitk/Build/MITK/mitkIncludes.cmake:80:
>  INCLUDE called with wrong number of arguments.  Include only takes one
> file.
>  Current CMake stack:
>  /cygdrive/c/MITK-cvs/mitk/Build/MITK/CMakeLists.txt;/cygdrive/c/MITK-cvs/mitk/Build/MITK/main.cmake;/cygdrive/c/MITK-cvs/mitk/Build/MITK/mitk.cmake;/cygdri
>  ve/c/MITK-cvs/mitk/Build/MITK/mitkIncludes.cmake
>
>  CMake Error: Error in cmake code at
>  /cygdrive/c/MITK-cvs/mitk/Build/MITK/mitkIncludes.cmake:98:
>  INCLUDE called with wrong number of arguments.  Include only takes one
> file.
>  Current CMake stack:
>  /cygdrive/c/MITK-cvs/mitk/Build/MITK/CMakeLists.txt;/cygdrive/c/MITK-cvs/mitk/Build/MITK/main.cmake;/cygdrive/c/MITK-cvs/mitk/Build/MITK/mitk.cmake;/cygdri
>  ve/c/MITK-cvs/mitk/Build/MITK/mitkIncludes.cmake
>
>
> -------------
> Settings
> -------------
> BUILD_FLTK_EXAMPLE               OFF
> BUILD_MITK_QCHILI_PLUGIN         OFF
> BUILD_QMITK                      OFF
> BUILD_QT_DESIGNER_PLUGINS        OFF
> BUILD_QT_EXAMPLE                 OFF
> BUILD_TESTING                    OFF
> BUILD_TUTORIALQMITK              OFF
> CMAKE_BACKWARDS_COMPATIBILITY    2.4
> CMAKE_BUILD_TYPE
> CMAKE_INSTALL_PREFIX             /usr/local
> DART_ROOT                        DART_ROOT-NOTFOUND
> DESIRED_QT_VERSION               4
> EXTERNAL_FUNCTIONALITIES_CMAKE
> GENERATE_DOCUMENTATION           OFF
> ITK_DIR                          /cygdrive/c/ITK-cvs/Insight
> MITK_CONFIG
> /cygdrive/c/MITK-cvs/mitk/Build/MITK/MITKConfig.cmake
> QMITKDATAMANAGER                 OFF
> QMITKDATATREEFILTERDEMO          OFF
> QMITKIMAGECROPPER                OFF
> QMITKISOSURFACE                  OFF
> QMITKLEVELSETS                   OFF
> QMITKMEASUREMENT                 OFF
> QMITKMOVIEMAKER                  OFF
> QMITKSIMPLEEXAMPLE               OFF
> QMITKVIEWINITIALIZATION          OFF
> QMITKVOLUMETRY                   OFF
> QMITKVOLUMEVISUALIZATION         OFF
> QT_QMAKE_EXECUTABLE              /cygdrive/c/Qt/4.2.3/bin/qmake.exe
> QT_QT_LIBRARY                    /usr/lib/qt3/lib/libqt-mt.dll.a
> USEDARKPALETTE                   OFF
> USE_KITWARE_DARTBOARD            OFF
> VTK_DIR                          /cygdrive/c/VTK-5/vtk-5.0.3

"Mark Alexiuk" hat am 3. Mai 2007 um 17:18 geschrieben:
> Hi,
> 
> I would appreciate any help on the following cmake error during a MITK build attempt.
> 
> The build MITK (0.6) uses cygwin (and gcc 3.4.4) on Windows XP. 
> Since it's an out-of-source build, I run ccmake from the build directory. 
> I've turned off all the cmake settings I can think of (see listing below).
 
Hi Mark,

I'm not absolutely sure on this, but I think that older versions of MITK did not support user-driven out-of-source builds (Nevertheless, MITK will be build in its own subtree, so you eventually will get your out-of-source build anyway ;) ). In addition, I would strongly advise you to use a more recent version of MITK if you can. 0.6 is nearly a year old and contained a bunch of nasty bugs and inconveniences which were corrected during the past year...

Regards
 Hendrik

-------------------------------------------------
Hendrik Belitz,
 Computer Scientist

EMail: hbelitz@... or hbelitz@...