[vtkusers] Using a boxwidgets transform to rotate a non-vtk 3D model?

[Totte Karlsson]

On 8/6/2015 7:40 AM, Cory Quammen wrote:
> TK,
>
> You can get the orientation in angle-axis representation with
>
> double wxyz[4];
> t->GetOrientationWXYZ(wxyz);
>
> From the docs [1]:
>
> The angle (w) is in degrees and the axis (xyz) is a unit vector.
>
> You should be able to apply this rotation to your molecule model's 
> RotateAboutVector function.
>
>
Thanks for the response.
I modified the boxwidget example in order to test your approach using a 
vtk actor, instead of my model, and what I believe I'm seeing is that 
the data from getOrientationWXYZ () is from the 'initial' orientation, 
instead of a progressive transformation. E.g using the following code in 
the callback function:

   virtual void Execute(vtkObject *caller, unsigned long, void*)
     {
         static int callCount = 0;
         vtkSmartPointer<vtkTransform> t = 
vtkSmartPointer<vtkTransform>::New();
         vtkBoxWidget *widget = reinterpret_cast<vtkBoxWidget*>(caller);
         widget->GetTransform(t);
         //widget->GetProp3D()->SetUserTransform(t);

         double* wxyz = t->GetOrientationWXYZ();
         if(anActor)
         {
             cout<<"CallCount: "<<callCount<<" Angle "<<wxyz[0]<<endl;
             anActor->RotateWXYZ(wxyz[0], wxyz[1], wxyz[2],wxyz[3]);
         }
         callCount++;
     }

Gives an angle output like this:
CallCount: 0 Angle 0
CallCount: 1 Angle 0.848528
CallCount: 2 Angle 1.69706
CallCount: 3 Angle 2.54558
CallCount: 4 Angle 3.39411
CallCount: 5 Angle 5.9397
CallCount: 6 Angle 6.78823
CallCount: 7 Angle 7.63675
CallCount: 8 Angle 8.48528
CallCount: 9 Angle 8.48528
CallCount: 10 Angle 8.48528

when moving the widget slightly with the mouse. Each 'callcount' 
represent a movement using the mouse.

Question is, how would one go about getting the 'latest' transformation, 
instead of the 'accumulated' one from the box widgets GetTransform()?

I guess I could save the transformation each time and do a 'back 
transformation' before applying the new transformation. But that would 
be more work (I assume), than applying an  incremental rotation.
-tk




>
> [1] 
> http://www.vtk.org/doc/nightly/html/classvtkTransform.html#aa8244cbab95a2dbb20e94af6e7f16b7f
>
>
> On Wed, Aug 5, 2015 at 8:54 PM, Totte Karlsson 
> <totte at dunescientific.com <mailto:totte at dunescientific.com>> wrote:
>
>     Hello,
>     I'm using vtk to visualize 3D models of molecules.
>     The actual molecular model is decoupled from vtk code, and
>     visualization is implemented using a model-view kind of design.
>
>     Currently I have the following challenge. If a user wants to
>     manipulate the absolute orientation of the molecule, it is done by
>     manipulating the underlying model using rotation functions, like
>     RotateX, RotateY, RotateAboutVector, and translation functions
>     TranslateX, Y etc.
>
>     For now 3D rotations are simply done using two scrollbars, one
>     horizontal and one vertical, causing the underlying model being
>     rotated about the vtk viewUP vector and the vtk cameraPlane normal
>     respectively.  This works fine, but is not so intuitive for the user.
>
>     Instead I believe allowing interaction with something like a box
>     widget that encapsulates the molecule view, (with disabled
>     handles), would be better.
>
>     The challenge is how to figure out how to capture the boxwidgets
>     orientation in a callback, and using this information in order to
>     manipulate the underlying molecular model.
>
>     For example, a boxwidget callback (from boxwidget example at
>     http://www.vtk.org/Wiki/VTK/Examples/Cxx/Widgets/BoxWidget):
>
>     class vtkMyCallback : public vtkCommand
>     {
>         public:
>             static vtkMyCallback *New()       { return new
>     vtkMyCallback;    }
>
>         virtual void Execute(vtkObject *caller, unsigned long, void*)
>         {
>             vtkSmartPointer<vtkTransform> t =
>     vtkSmartPointer<vtkTransform>::New();
>             vtkBoxWidget *widget =
>     reinterpret_cast<vtkBoxWidget*>(caller);
>             widget->GetTransform(t);
>
>             double* angles?? = t->GetOrientation();
>             if(aMoleculeModel)
>             {
>                 //Use some secret(!) code to manipulate orientation of
>     a molecule to 'rotate' it together with the boxWidget
>                 aMoleculeModel->RotateZ(??);
>                 aMoleculeModel->RotateY(??);
>                 aMoleculeModel->RotateZ(??);
>             }
>         }
>         Molecule* aMoleculeModel;
>
>     Anyone seeing how to do this or have some pointers?
>
>     Cheers,
>     tk
>
>     _______________________________________________
>     Powered by www.kitware.com <http://www.kitware.com>
>
>     Visit other Kitware open-source projects at
>     http://www.kitware.com/opensource/opensource.html
>
>     Please keep messages on-topic and check the VTK FAQ at:
>     http://www.vtk.org/Wiki/VTK_FAQ
>
>     Search the list archives at: http://markmail.org/search/?q=vtkusers
>
>     Follow this link to subscribe/unsubscribe:
>     http://public.kitware.com/mailman/listinfo/vtkusers
>
>
>
>
> -- 
> Cory Quammen
> R&D Engineer
> Kitware, Inc.

-- 
------------------------------------------------------------------------
<http://www.dunescientific.com/> 				
Totte Karlsson: totte at dunescientific.com 
<mailto:totte at dunescientific.com> or (425) 780-9648 
<tel:%20425%20%20780-9648>
http://www.dunescientific.com/
© 2015 Dune Scientific, LLC. All rights reserved.



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/vtkusers/attachments/20150806/7110bfdf/attachment.html>