This issue has been considered in a version before every tool was updated to the new interaction concept. As far as I remember any interaction at another slice than the one you started the live wire tool at was rejected.

We've tried to discover this issue during bugsqashing but were not able to reproduce it. Can you provide us with some more hints?

I'm not able to reproduce this bug on my machine.
Also checked it on another machine running MacOSX 10.9 with same configuration.

Needs further investigation. Can not be fixed within current release cycle.

[21904d]: Merge branch 'bug-17375-3D-renderwindow-interaction-tests'

Merged commits:

2014-03-12 09:11:22 Tobias Schroeder [61b564]
Disconnect all observers.

CanHandleEvent still exists in mitkSegTool2D and checks the mapperID. This might be used to prevent interaction in 3D.

Changes of T16780 fixed the problem.

[8b0cc6]: Merge branch 'bug-17341-fix-LW-tool'

Merged commits:

2014-03-26 16:08:21 Tobias Schroeder [b985ef]
Alternate user feedback contour to be corner based

I've adapted the user feedback contour to be corner based thus the shift of the contour when filled in the slice is removed. It is not the best solution since some pixels marked by the algorithm are not taken into account but this way the user isn't confused about what is happening when confirm segmentation button is pushed.

Since it is a container for mitkContourModel the FillContourInSlice function of mitkContourModelUtils can be used.

The problem seems to be fixed taking commented out line "pol2stenc->SetTolerance(0);" in FillContourInSlice back in.
This line is also present in the fill polydata contour example from VTK's homepage.

[ee7516]: Merge branch 'bug-17340-PaintTool-missing-feedback-contour'

Merged commits:

2014-02-26 15:03:17 Tobias Schroeder [2454ee]
Fixed missing feedback contour

Workaround behaves like ipSegmentation.

Problem is related to center based contour output of ImageLiveWireFilter. The FillContourInSlice seem not working correctly with contours with point coordinates in pixel center.

Added a transition to statemachine PressMoveReleaseWithCTRLInversionAllMouseMoves for mouse move. Adapted callback functions in PaintBrushTool to handle left mouse button pressed during mouse move.

Use MITK_ACCESSBYITK_DIMENSIONS with 2,3,4

Adapted classes according to code review.

[5332ae]: Merge branch 'bug-16871-fix'

Merged commits:

2014-03-05 13:52:06 Tobias Schroeder [d43595]
Module was renamed

[30c0e8]: Merge branch 'bug-16871-Interaction-Testing-Framework-integration'

Merged commits:

2014-02-27 11:25:34 Tobias Schroeder [052e59]
Adapted to new API of InteractionPositionEvent.

[5d6013]: Merge branch 'bug-16871-fix'

Merged commits:

2014-02-27 15:51:31 Tobias Schroeder [672167]
COMP: Use cmake variable MITK_ENABLE_RENDERING_TESTING to enable interaction test. Avoid crash on linux due to vtkRenderWindow creation.

[939460]: Merge branch 'bug-16871-fix'

Merged commits:

2014-02-27 14:42:33 Tobias Schroeder [4656a5]
COMP: mitkFiberfoxAddArtifactsToDwiTest used wrong data because of MITK-Data update.

Reopen bug due to CMake output in Core/Code/Testing.

Added an InteractionTestHelper class and an example for PointSetDataInteractor of how a test based on CPPUnit will look like.

• Added config section to interaction file for renderer names and default view directions to dynamically create render-windows in helper class.
• Support for KeyPressEvent.
• Setter for timeStep of all render-windows.

[accf5d]: Merge branch 'bug-16871-fix'

Merged commits:

2014-02-27 12:49:16 Tobias Schroeder [b8f1a7]
COMP: fix windows compile errors.

Did some additional stuff to get Tests working in all render windows of mitkWorkbench by providing three render windows with default view directions.
Also registered a DisplayInteractor to handle scroll, move and zoom events.
Added function to InteractionTestHelper to set timestep for 4D data support.
Unregister Renderer instances in destructor.

Managed to playback scroll interaction the right way. There was a property missing in the xml file for ScrollDirection.

I've fixed several things in the plugin.

We found out that the inversion of the direction was due to an implicit conversion of the axis of displayPosition in the constructor of InteractionPositionEvent. Responsible for that was BaseRender::Map2DRendererPositionTo3DWorldPosition where GetDisplayGeometry()->ULDisplayToDisplay was used with the same variable for both parameters. This caused the y-axis to be flipped again when an event was created from the xml file.

[774456]: Merge branch 'bug-16687-tool-testing-integration2'

Merged commits:

2014-03-13 09:18:59 Tobias Schroeder [4d3ea1]
Adapted use of MITK_ASSERT_EQUAL

New remote branch pushed: bug-16690-segmentation-picking-tool

[d5c8c8]: Merge branch 'bug-16690-segmentation-picking-tool'

Merged commits:

2014-01-17 10:20:36 Tobias Schroeder [a60f5a]
Added ui file for PickingTool.

New remote branch pushed: bug-16687-tool-testing-framework

Renamed bug and added dependency T16871

[5f8725]: Merge branch 'bug-16651-segmentation-statemachine-fix'

Merged commits:

2014-02-26 14:32:35 Tobias Schroeder [6ca78e]
Added missing statemachine for tool with wheel interaction.

[b4d5f2]: Merge branch 'bug-16651-tools-have-been-adapted-to-new-interaction-fra

Merged commits:

2014-02-24 13:43:59 Tobias Schroeder [7c724f]
Fix delete point in LiveWire tool

Class diagram

Class diagram

[b9a546]: Merge branch 'bug-16493-ContourModel-Recursion-fix'

Merged commits:

2013-11-25 13:48:42 Tobias Schroeder [8c0846]
Fixed recursive call leading to infinite loop.

Design approach

New remote branch pushed: bug-16493-ContourModel-Recursion-fix

I will integrate your fix as soon as possible.

Solution looks fine.

[7850cf]: Merge branch 'bug-16336-ContourModel-rendering-ignores-opacity'

Merged commits:

2013-10-28 16:33:59 Tobias Schroeder [a11760]
Set default opacity value.

[8fe4c8]: Merge branch 'bug-16302-fix'

Merged commits:

2013-11-11 15:31:44 Tobias Schroeder [858537]
COMP: fix test.

[27e3b7]: Merge branch 'bug-16302-fix'

Merged commits:

2013-11-11 14:59:51 Tobias Schroeder [51dab0]
COMP: fix test.

[1fa18c]: Merge branch 'bug-16302-fix'

Merged commits:

2013-11-11 14:29:57 Tobias Schroeder [62cf07]
COMP: fix unresolved test function.

[f794c2]: Merge branch 'bug-16302-ContourModel-API-improvements'

Merged commits:

2013-11-11 13:42:52 Tobias Schroeder [2acc1c]
Renamed funcion to SetClosed(bool) in ContourElement.

New remote branch pushed: bug-16302-ContourModel-API-improvements

[0a1296]: Merge branch 'bug-16302-fix'

Merged commits:

2013-11-11 14:17:38 Tobias Schroeder [2c3222]
COMP: fix compile error on windows and corrected test condition.

I'm currently not able to participate on bugsquashing. But the background colour quickfix should be easy. I'll find someone for next bsq.

Yes. I've just removed the line in SegTool2D where the operations where added to the stack.

In an undo/redo operation the segmentation image is stored only as a pointer but the current slice is copied. This definitely increases memory. There are some lines in the constructor of the operation (mitkDiffSliceOperation - line 52) commented out that used a zip container for the copied slice. A first approach might be to compress the undo/redo image slices again. As the slices are only containing values of 0 and 1 a zip compression should be very effective.

Fixed with 16004

Crash is due to adding points while updating the itk filters. Can be fixed by removing the pointset interactor during update steps and adding it again afterwards.

after line 286 in mitkFastMarchingTool3D.cpp

New remote branch pushed: bug-16097-crash-in-fastmarching-tool-quickly-adding-points

I'm not able to reproduce this bug on Mac OS X (current master).
It occurred on my Win 7 64bit machine.

pointset

pointset

no_name.nrrd

Crosshair color has gone

no_name.nrrd

New remote branch pushed: bug-16067-disable-contour-interaction

We need to adapt the segmentation test lists according to LiveWireTool. The list should contain the hint that the contours will not be able to be edited after deactivate the tool.

[f1127e]: Merge branch 'bug-16067-disable-contour-interaction'

Merged commits:

2013-09-12 09:28:21 Tobias Schroeder [43da26]
Don't remove contours from DataStorage.

The reason for this bug might be the size/origin/extent of the current world-geometry which is used to extract the slice of the working image. In case of a global bounding geometry the size exeeds the geometry of the reference image.
If a 'reinit' is performed on the current reference image the ContourTool is working as expected. The strange behavior is present when a new segmentation is created and after a 'global reinit'.
Even more annoying is the fact that the view stays the same after creating a segmentation on a different segmentation image (with 'reinit' performed on current). Thus no contours are ever visible because the renderwindow only shows the previous reference image.
A simple 'reinit' might fixed this issue for the user but will not resolve the bug itself.

As far as I remember we had that reinit in the segmentation plugin before the UI redesign.
I had a look at git commit history:
The reinit was gone with commit 7944e892 "General code cleaning. Removed a lot of functions which are obsolet or unnecessary by now."
ForceDisplayPreferencesUponAllImages() was deleted. In this function the actual reinit was done by
// Reinit current node
mitk::RenderingManager::GetInstance()->InitializeViews(
referenceData->GetData()->GetTimeSlicedGeometry(),

mitk::RenderingManager::REQUEST_UPDATE_ALL, true );

Does anyone know why this code was deleted? Was there an actual reason or had it been just accidentally removed with other code.

New remote branch pushed: bug-16063-2D-segmentation-is-buggy-with-more-than-one-image

The crash is caused by alternating the parameters on start up. In case no seed point has been set yet the itk filter is still NULL.

New remote branch pushed: bug-16047-crash-in-fastmarching-tool-ui

[af00b3]: Merge branch 'bug-16047-crash-in-fastmarching-tool-ui'

Merged commits:

2013-09-11 17:10:45 Tobias Schroeder [8e1f3e]
Fixed access to NULL pointer on changing parameters.