[vtkusers] Strange behavior in vtkPlaneSource (maybe bug?)

[Marco Sambin]

Hi Berk,

thank you for your message and explanation, which makes perfect sense.

However, in this case I believe there may be an issue in the source code of
the vtkImagePlaneWidget class. In particular, in the method
vtkImagePlaneWidget::OnMouseMove(), I see:

void vtkImagePlaneWidget::OnMouseMove()
{
[...]
    else if ( this->State == vtkImagePlaneWidget::Rotating )
    {
    camera->GetViewPlaneNormal(vpn);
    this->Rotate(prevPickPoint, pickPoint, vpn);
    this->UpdatePlane();
    this->UpdateMargins();
    this->BuildRepresentation();
    }
[...]
}

The Rotate() method, in turn, ends with the following calls:

void vtkImagePlaneWidget::Rotate(double *p1, double *p2, double *vpn)
{
[...]
  double newpt[3];
  this->Transform->TransformPoint(this->PlaneSource->GetPoint1(),newpt);
  this->PlaneSource->SetPoint1(newpt);
  this->Transform->TransformPoint(this->PlaneSource->GetPoint2(),newpt);
  this->PlaneSource->SetPoint2(newpt);
  this->Transform->TransformPoint(this->PlaneSource->GetOrigin(),newpt);
  this->PlaneSource->SetOrigin(newpt);
}

Hence, the Point1, Point2, and Origin of the vtkPlaneSource are changed in
this order.
Now, the OnMouseMove() method calls UpdatePlane() immediately after
Rotate(), and UpdatePlane in turn calls GetNormal() on PlaneSource without
calling any Update() on it first:

void vtkImagePlaneWidget::UpdatePlane()
{
[...] // No call to "this->PlaneSource->Update()" is present in the omitted
code...

  double normal[3];
  this->PlaneSource->GetNormal(normal); // At this point the retrieved
normal may be incorrect
[...]
}

Maybe a "this->PlaneSource->Update()" call is missing somewhere?
Thanks again for your comments.

Best regards,

Marco Sambin


On Tue, Apr 2, 2013 at 5:19 PM, Berk Geveci <berk.geveci at kitware.com> wrote:

> Please note that this is not a utility class. It is an algorithm. ALL
> output that matters is produced after Update(). The fact that the normal is
> updated in some cases before Update() is coincidental and you should not
> rely on this behavior. If you are interested in calculating the normal of a
> plane given 2 vectors, I suggest using a cross product as provided by
> vtkMath.
>
> Best,
> -berk
>
>
>
> On Thu, Mar 28, 2013 at 7:15 AM, Marco Sambin <m.sambin at gmail.com> wrote:
>
>> Hi all.
>>
>> I am using the vtkPlaneSource class in a VTK-derived class. I noticed
>> that the following simple sequence of calls:
>>
>> =========================
>> CODE A:
>>
>> double newO[3];
>> double newP1[3];
>> double newP2[3];
>> double newNormal[3];
>>
>> // [...] set new values...
>>
>> myPlaneSource->SetPoint1(newP1);
>> myPlaneSource->SetPoint2(newP2);
>> myPlaneSource->SetOrigin(newO);
>> myPlaneSource->GetNormal(newNormal);
>> =========================
>>
>> finally returns a normal which seems INCORRECT to me.
>> On the other side, the following sequence of calls (which differs from
>> CODE A in the position of the SetOrigin() call):
>>
>> =========================
>> CODE B:
>>
>> double newO[3];
>> double newP1[3];
>> double newP2[3];
>> double newNormal[3];
>>
>> // [...] set new values...
>>
>> myPlaneSource->SetOrigin(newO);
>> myPlaneSource->SetPoint1(newP1);
>> myPlaneSource->SetPoint2(newP2);
>> myPlaneSource->GetNormal(newNormal);
>> =========================
>>
>> finally returns a normal which seems CORRECT.
>> After taking a look at vtkPlaneSource.cxx source, I believe that the
>> problem is that while the implementation of the SetPoint1() and SetPoint2()
>> methods finally calls UpdatePlane(), which also updates the normal, the
>> implementation of SetOrigin() does not update the normal. Hence, if
>> SetOrigin() is the last call, the resulting normal is incorrect in my
>> opinion.
>>
>> Can you please comment?
>> Thanks in advance,
>>
>> Marco
>>
>>
>> _______________________________________________
>> Powered by 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
>>
>> Follow this link to subscribe/unsubscribe:
>> http://www.vtk.org/mailman/listinfo/vtkusers
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.vtk.org/pipermail/vtkusers/attachments/20130405/e9ed85bd/attachment.htm>