There is no longer a use case for this kind of segmentations. If this issue reoccures please contact Sebastian Schäfer or Markus Engel.

[620cdc]: Merge branch 'bug-11106-gantry-tilt-cleaned-up'

Merged commits:

2012-08-27 14:17:58 Johannes Kast [b88875]
changed mitk data revision for new tilted test data

Daniel and I finished the work on this feature and added additional tests to the DICOMLoading Tests.

This change was only pushed to the branch releases/snapshot-201109-2 until a more general solution for this problem is provided.

Meanwhile MITK has a more advanced LUT implementation. Its safe to close this ticket.

The unit test of mitkLookupTable was extended to use all (and future) lookup tables in the proposed provider class. This is also part of the patch. The test was not started and compiled yet.

Implementation of LUT Provider including Test

Implementation of LUT Provider including Test

Sorry, too late for the last comment...

Bei der Übergabe der map aus dem "scanner" objekt (->GetMappings()) wurde die Map zwar kopiert aber die Inhalte der Map Schlüssel (char Pointer) wurden gelöscht. In einer Debug Umgebung unter Windows (Inhalte hinter Pointern werden bei Aufruf des destruktors überschrieben) kommt es dann zum Crash beim Zugriff.

Push mit Hash: d403326d36fa9f0292ee5bf2ef905b503292285f

[d40332]: Merge branch 'bug-7339-fix-wrong-dicom-property-retrieve'

Bitte um Review der Änderungen.

Commit in T5164 . REV: 28902

This is fixed. Commit follows ...

[SVN revision 27193]
FIX (#5874): added additional check for memory size

[SVN revision 26649]
ENH (#5682): added color property for the volume annotation; added string property for the unit of the volume annotation;

[SVN revision 26653]
COMP (#5682): removed unused variable;

Kann geschlossen werden...

The unit should be also customizable. Siemens uses in his Syngo application cm3 instead of ml.

We should definitly write a unit test for this. This is still relevant.

Hi, I do not know how much work this is but at the moment it is not possible for me to free some time for IGT... Sorry Michi -> assign goes back

Fixes linker error in mitkCore caused by a changed interface in VTK IO

The release version of VTK 5.6.0 has a changed interface in vtkImageReader::SetDataMask().

Last comment (SVN) has a wrong Bug ID: The comment is related to T3308

[SVN revision 21370]
ENH (#3341): added PropertyList to RenderingManager

Implementation is finished. I created the unit test for the rendering manager see T1985

[SVN revision 21371]
ADD (#3341): added test for RenderingManager for testing the PropertyList see also T1985

@Sascha
On which machines does the test fail. For us on Windows 7 64bit everything went fine so far.

Data is everything ;-)

As this is the new job of Markus Engel (refactoring MultiWidget/render windows) this is in my opinion not 1.0 relevant.

Meanwhile the design of the of this widget has changed several times this is no longer an issue.

The core dependency issue is resolved. The missing support for drag and drop in the ExtApp is moved to T3377

No progress yet. Still fighting with updates and patches...

I suggest a rename of the SceneReaderInterface to SceneFactoryInterface since the SceneIO class is also capeable of writing whole scenes.

This is 1.0 relevant... I will try to fix it.

[SVN revision 20318]
CHG (#3003): added a method for cleaning up the segmentation filter

As this bug affects only shape model segmentation tools this is not 1.0 relevant.

We can close this one. Tobi do you agree?

Talked to Tobi he agreed to close this one

In release mode it is not crashing on my 64bit windows. "Abort Segmentation" is crashing.

I suggest to use this models for the installer:

It seems that the slice interactor is shut down with the start of the painting tools. Is there a way to suppress this?

[SVN revision 20116]
CHG (#2950): enabled drag and drop support for *.mitk scene files

I think this behavior is caused on the selection service.
Deselected reference node -> ToolManager gets notified -> there is no reference image anymore for the tool -> tool is deactivated -> ...

I will try to fix the DnD loading today.

[SVN revision 19926]
ADD (#2864): added menu entry File->Save Project... as a FileSaveProjectAction. The scene can now be stored with this mechanism; added .mitk file extension to the object factories; enabled load mechanism for .mitk files over the menu

[SVN revision 20283]
CHG (#2699): included the new icons

The icons in the segmentation folder are to small. Perhaps we try larger ones. Marco can you just exchange the pictures with the right names and identifiers in the images?

Three different simple themes

What you need are the files as mentioned in commend #2 from Sascha. I played around with the Qt paletts (dark, shadow, light, etc. see Qt Doku) but for a really nice skin you have to use own colours for the different widgets.

Three different simple themes

Around line 222 in mitk::VolumeVisualizationImageProcessor (rev. 18958) a OrImageFilter is used just to create a copy of the input image. This should be changed to a nicer way of copying an image.

The refactoring is done. See the change request file for the made changes.

ChangeRequest

Change Request

Change Request

Now we can use mouse wheel interactions. A first manual test of the wheel movement information inside of a bundle was positive.

In mitkEventTypeDef.h you find "typedef DisplayPositionEvent WheelEvent;".
This works because by now everyone uses the WheelEvent for clicks and the resulting position. Now we want additional information about the mouse wheel movement. So we should implement a new class mitkWheelEvent derived from DisplayPositionEvent which has an additional parameter for the mouse wheel movement (delta).
Also the QmitkEventAdapter class has to be modified to throw proper mouse wheel events.

ChangeRequest

The bug needs to be reopend due to neccessary changes in QmitkEventAdapter

ChangeRequest

A first version was commited. There are still some Qt related signal/slot issues...

The signal slot problems are solved. The methods inside the old ui.h file moved to the QmitkIGTExampleView class.
Due to Qt4 porting the old state (PropertyList) of the bundle is not stored anymore.

I am closing this as the test coverage is good enough.

The SceneIO test now has a coverage with 69%. For further improvements make checks for empty and already filled datastorage and also save render specific properties.

@Daniel is there a bug for the scene serialization not storing BaseDatas ?

Is there any reason why the PropertySerializationTest is not in the configuration? It works fine for me on Windows 7 64.

This was not a bug. In building up the data storage was a c&p error...

I will write a test the next days...

The reading part of the SceneIO test revealed that some data is not stored:

A test was added for the error calculation. Two pointsets are created with two cubicals. Those are registered and the error calculations of the filter are compared to the known error values.

First progress from the bug squashing party for BUG 2132

First progress from the bug squashing party for T2132

Also loading a PointSet and interpolate between those points would be nice.

Was implemented. Following additional functions are now available for gathering registration errors:

ScalarType GetFRE() const;
ScalarType GetFREStdDev() const;
ScalarType GetRMSError() const;
ScalarType GetMinError() const;
ScalarType GetMaxError() const;
ScalarType GetAbsMaxError() const;

The bundle was ported but there is still work to do:

This is already fixed. Bug closed...

This is 1.0 relevant. I will check if there is a new implementation. If not I will use the preference store to resolve this issue.

Is now a party bug

A test was submitted...

All features are implemented