kislinsk added a revision to
T29324: Create view for pixel information:
Unknown Object (Differential Revision).
Reset Classifier available
Update QmitkInteractiveFiberDissectionView.cpp
Pushed new branch to rMITK MITK: .
Deleted branch from rMITK MITK: .
Make tips on Welcome page optional
Merge branch 'feature/T29376-MakeTipsOptional' into develop
Pushed new branch to rMITK MITK: .
Deleted branch from rMITK MITK: .
Use a more adequate initial workbench window size
Center initial window to screen
Adapt perspective ratios to 16:9 workbench windows
Show views before perspectives
Merge branch 'bugfix/T29348-ImproveWelcomePage' into develop
Update mitkStreamlineInteractorBrush.cpp, mitkStreamlineInteractorBrush.h, and…
Update mitkStreamlineFeatureExtractor.cpp, files.cmake, and 4 more files...
Update files.cmake.autosave
Update mitkStreamlineFeatureExtractor.cpp, QmitkInteractiveFiberDissectionView.
Remove Certain negative data from Testdata
Update .gitignore, mitkStreamlineInteractor.cpp, and 2 more files...
adding visualisatzion of most certain data
Bug fixed regarding Prototypes
Shuffle, when Resampling a Streamline. Validation added
Experiments with fixed number of Prototypes
Update mitkFiberBundle.cpp, mitkStreamlineInteractor.cpp, and 6 more files...
Update mitkFiberBundle.cpp, mitkFiberBundle.h, and mitkStreamlineInteractor.cpp
FFT for in uncertainty range implemented
Update mitkFiberBundle.cpp, mitkFiberBundle.h, and 8 more files...
SFF Prototype selection implemented
Update mitkStreamlineFeatureExtractor.cpp, mitkStreamlineFeatureExtractor.h…
Interacotor: Unlabel data
Change number of Resampling points to 40
Random Forest implemented with uncertainty labeling...
Features are now saved, which reduces computing time.
Creation of uncertainty labes done
Back up. Random Forest makes Prediction, but its bad
First iteration of cycle implemented. Prediction is done on both training and…
Init of calculation D_mdf which is now possible between (hardcoded) Prototypes…
Ready for Prediction, but test array is to big. Need to change type op cv::Mat
Init of StreamlineInteractor
Merge branch 'CreateModule' into InteractiveFiberDissection
First steps for new Module
Init of DataInteractor for StreamlineLabel and FeatureExtractor
Init of Interactive FiberDissection
Pushed new branch to rMITK MITK: .
kislinsk renamed
T29375: Cursor offset in 2D tools with Qt 5.15 on macOS from
[Segmentation] Cursor offset in 2D tools with Qt 5.12 on macOS to
Cursor offset in 2D tools with Qt 5.12 on macOS.
This bug is originating in Qt 5.15. It is reproducible but not really related to something in segmentation. It's just an obvious part of MITK to encounter the offset.
kislinsk renamed
T29375: Cursor offset in 2D tools with Qt 5.15 on macOS from
[Segmentation] Cursor offset in 2D tools to
[Segmentation] Cursor offset in 2D tools with Qt 5.12 on macOS.
Ya, I tried. Both before and after annotation. Changes nothing. Bug exists.
Could you explicitly "reinit" the reference node and see if that changes anything?
Bug likely to be caused by Qt 5.15 in my local.
This may not affect this release build since official release builds use Qt 5.12.
Reworked the tools to make them faster.
Close ticket for now. If It is not fast enough we need to make more rework and need to make a in depth profiling where we loose the time.
Remove new reinit overlay for StdMultiWidget
Remove unnecessary function overriding
Remove automatic state definition
Add geometry comparison if coupled render windows are found
Add virtual function for subclass
React only if coupled render windows are found
Provide function to check for coupled render windows
T29359 - Fixed inversion feature. Now wipe and paint pixels are indicated at…
Fixed wrong active label queries
Fixed Undo/Redo behavior of paintbrush tools
Fixed inversion feature. Visual representation is correctly showing the mode…
I implemented the feature based on the shift key:
Deleted branch from rMITK MITK: .
Optionally skip done tasks
Merge branch 'feature/T29372-OptionallySkipDoneTasks' into develop
Pushed new branch to rMITK MITK: .
There are valid points. We should not confuse the user.
Correctly done, I think the new overlay style is the way to go. Besides optimized/userfriendly text (+making the overlay blocky), we could also directly add icon and btns at the overlay to allow reinit of the render window or all render windows. I think from a UX perspective it would be a huge improvment compared to all other versions.
Deleted branch from rMITK MITK: .
Improve standard display icon
Improve all standard display icons
Merge branch 'bugfix/T29373-ThemeStdDisplayIcon' into develop