(In reply to Alfred Franz from comment #1)

We have to check if the mitkItkNonUniformBSpline contains patented code,
which might be the reason why it was removed from ITK. If yes we have to
remove it as soon as possible and this bug has very high priority.

If not, we can leave it as it is for the moment and remove this class later.

However: this class should be declared deprecated either way!

any comments on this?

lokale Codeänderung ohne Nebenwirkungen. Fehler tritt nicht mehr auf.
mitk::PlanarFigure::AddControlPoint() ist ziemlich komplex verschachtelt, das könnte man sicher bei Gelegenheit mal vereinfachen.

Fehler ist behoben, tritt nun nicht mehr auf. Codeänderung ist klar und einfach, alles okay.

is this merged into the snapshot-201105-preparation branch? Doesn't look like it.

Take a look at http://www.vtk.org/doc/nightly/html/classvtkCameraWidget.html. This implements a similar behaviour.

now, it uses the standard text color. If the user manually changes his console background/text colors, it will still use them.

[SVN revision 28876]
CHG (#6521): print log messages uncolored

re-assign to MacOS specialist.

@Daniel, can you look into this?

This predicate can be used to check, if a node is a parent to a given node. It can be computationally quite expensive for a large DataStorage.

you name it, you have it.

After my last update, I get a lot of warnings on the console window that could have something to do with your changes:

Bastian, please find a suitable assignee for this bug, I won't have time to work in this.

[SVN revision 23276]
FIX (#4179): add segmentation as child node of source image

We don't have the necessary hardware to test it. It should theoretically work with the same code. After all, it is not "Aurora" code, just "NDI" code and the API is the same for both systems. We don't need to put much effort into this, but we should keep it in mind for the future or for external users of MITK-IGT.

Just a small remark: There are hybrid Polaris systems that use wired tools. These tools probably have their rom file 'onboard' and can be discovered using the NDI API, just like the wired Aurora tools. With the current code, the discover-wired-tools button is hidden for all Polaris systems, not ust the passive systems.

• Changes in CoreHelpers.cmake are sound and straightforward.
  • No further suggestions, setting status to VERIFIED.

Date: Mon Apr 19 14:28:49 2004 CEST (6 years ago)

ToolPairNavigation is Ingmars Bundle.

Wikipedia basically writes that they are synonyms. In the papers I read, they are used as synonyms too, but I read 'fiducials'/'fiducial markers' more frequently than 'landmarks'.

Fiducial <--> Landmark - aren't they synonyms?

After discussion in the MITK meeting, I used this solution:

Now there are a total of 8 commits in itkNonUniformBSpline.h and itkNonUniformBSpline.txx. Using the new version, MITK compiles with ITK 3.18 in Debug (and Release) mode.

[SVN revision 22710]
FIX (#3839): import and use CVS version of itkNonUniformBSpline class to circumvent a bug in ITK 3.18 (see http://public.kitware.com/Bug/view.php?id=10633)

The problem is actually a bug in ITK 3.18.

Free for all bugsquashing bug.

Party bug, can be done by anyone.

Could be, but it is strange, that the test is succeeding on most other windows and linux dartclients. See http://mbits/cdash/testSummary.php?project=3&name=mitkTrackingDeviceSourceTest&date=2010-04-22
and http://cdash.mitk.org/testSummary.php?project=1&name=mitkTrackingDeviceSourceTest&date=2010-04-22

Ingmar, I won't have time for this. Do you want to look into it? (Otherwise, reassign or just close the bug)

Party bug, can be done by anyone.

Paul, you are working in this area, please take this bug.

Party bug, can be done by anyone.

Party bug, can be done by anyone.

Paul, you are working in this area, please take this bug.

[SVN revision 22361]
FIX (#3726): prevent terminating the tracking thread before it has started.

can what?

Party bug, can be done by anyone.

Paul, you are working in this area, please take this bug.

Party bug, can be done by anyone.

refactoring mostly done. Interface changes are committed.

[SVN revision 21905]
ENH (#3454): add new unit test for VirtualTrackingDevice

Reset to default Assignee

[SVN revision 21911]
CHG (#3454): VirtualTrackingDevice uses VirtualTrackingTool now, removing obsolete friend declaration

• write unit test for mitk::VirtualTrackingDevice

[SVN revision 21904]
ENH (#3454): add new VirtualTrackingTool class

[SVN revision 21903]
FIX (#3454): store tool related data in new VirtualTrackingTool object, refactor VirtualTrackingDevice class accordingly, use mutexes

The ITK Spline class that we use is now deprecated:
http://public.kitware.com/Bug/view.php?id=10633#c26901 (instead of fixed...). When switching to new ITK versions, we might get problems with that.

The callback mechanism should be tested with a unit test too, not only with a manual test.
--> Core Request granted.

Reset to default Assignee

maybe we can use a tool like http://www.heise.de/software/download/imdisk/60650 to temporarily create a small virtual harddisk/ramdisk that we can fill up.

I'll fix it

Not this week, I can try to reproduce it next week.

[SVN revision 21592]
FIX (#3393): do not apply median filter to segmentation image

The code that changes the opacity is in QmitkSegmentationView:

the image is available at /home/jochen/temp/planungs-ct.pic

screenshot after cropping shows the cropped area lying to the left of the area that was selected. Note the fiducial point that was in the selected area but is outside the resulting cropped image

screenshot after cropping shows the cropped area lying to the left of the area that was selected. Note the fiducial point that was in the selected area but is outside the resulting cropped image

Cropping should update the geometry, so that the pixels of the cropped image are at the same position as before cropping. Everything else doesn't really make sense.
I'm pretty sure that ImageCropper behaved this way some time ago.

cropping bounding box in original image

Reset to default Assignee

[SVN revision 21421]
ADD (#3367): Add experimental IGT recorder/player view. Recording into XML files and pointsets works, replay feature is currently unfinished

Party bug. Can be done by anyone.

Adding test author to cc list. Alfred, please comment.

re-assigning to class author.

Core modification granted.

From the 'MITK is ExtApp' point of view, I agree.

It is a composite object that links the data object (BaseData) with associated
interactor, mappers and meta data (property lists). So from a pure
functionality view, it is not a node. Rather a connector object.

The new name for DataTreeNode should be discussed. CC'ing MITK developers.

setting component to 'segmentation'

If Daniel agrees, this can be done by anyone during the bug squashing party.

also, extend the filter to include tracking volumes of all supported tracking devices - also variants of existing devices like polaris spectra/vicra and MicronTracker S40, S60,...

TransformPosition and TransformOrientation will not be renamed. The old names are good enough.

I'm cleaning up my bug list, Alex, Ingmar, you should decide if this should be done and then reassign the bug to someone.

[SVN revision 21416]
CHG (#3271): rename SetBaseData() to SetRepresentationObject()

[SVN revision 21419]
COMP (#3271): rename SetBaseData() to SetRepresentationObject()