I created a segmentation of Pic3D, however, there is no option to chose DICOM SEG when saving the image. Might be related to the third incompatible commit? Is this really finished as I noticed includes of a cpp file which looks a little bit experimental in that last commit?
- Queries
- All Stories
- Search
- Advanced Search
- Transactions
- Transaction Logs
Advanced Search
Mar 11 2019
I cherry-picked the two compatible commits (out of three) into the releases/2018.04 branch.
Mar 8 2019
Mar 7 2019
Pushed new branch T24973-Bias.
Pushed new branch T25918-RenameStatisticsObject.
Mar 6 2019
Pushed new branch T25409-TractEnvDicomSeg.
Mar 5 2019
Was fixed "by the way". List entries have an "X" now.
Schould be done after T23808: Add Rule base class for DICOM source imagereferences
I tried it again with the current master using Ubuntu 18.04.
I still have no problems debugging into the mentioned functions.
If I add cpprestsdk|cpprest to the PACKAGE_DEPENDS of a module, the properties of the cpprestsdk::cpprest target are not respected, i.e. the module is not linked against the library, leading to unresolved symbol linker errors. However, if I do it in plain CMake, it works:
Mar 4 2019
Sure, here they are:
Fixed the dependencies to Boost and ZLIB. Changed the handling of zlib. The idea is, that it is not built if there is an easy to find system zlib. However, the decision is now made in the ZLIB script instead of the cpprestsdk script. Must be tested on Linux and macOS again, though.
Mar 1 2019
The crash is caused by the duplicate creation of the layers:
had a further look into it:
- looked into the matrix vectors. Looks like the first and the second vector is normalized, but the third is not (https://academo.org/demos/3d-vector-plotter/)
- tried to transpose the matrix
- tried to change the last column by sign
- the rectification leads to a normal vector that is the same like the first vector.
- we can set the matrix manually to space directions: (1.9,0,0) (0,1.9,0) (0,0,12) (pixel spacing). But that's not a real solution as we usually can't alter the data.
Pushed new branch T25831-improve-statistics-documentation.
Seems to be resolved... Can't reproduce it.
Pushed new branch T25854-reset-ui-after-image-change.
Builds now on Linux and macOS. However, the cpprestsdk::cpprest tarket does not link against C++ REST SDK. Have to investigate further. Also, our superbuild does not recognize the dependency to ZLIB.
Pushed new branch T26127-Bandwidth.
Pushed new branch T25845-GUIInitializationState.
Feb 28 2019
Pushed new branch T26121-Kspace.
Yes, Linux and macOS are next, but I currently have only access to a single computer, so it is a sequential process. :-)
Very nice!
Currently testing on Windows. Superbuild was successful.
Pushed new branch T26094-CppRestSdk.
Feb 27 2019
Pushed new branch T26117-UnusedTypedefs.
Feb 26 2019
Pushed new branch T26108-PartialFourierFix2.
Pushed new branch T25697-OptimizationOfMitkImageStatisticsCalculatorTest.
Feb 25 2019
Pushed new branch T26108-PartialFourierFix.
Pushed new branch T26106-FixMergeErrors.
Feb 22 2019
Pushed new branch T26105-FfInversion2.
Pushed new branch T26105-FfInversion.
ALso fix T1 calculation. Should be independent of readout.
Pushed new branch T26037-useCustomStatisticValues.
Feb 21 2019
Pushed new branch T26092-FfConstantSeed.
Test data: Fibers & DICOM T1 image.
Croshes on savin multi-label image as DICOM Seg.
Feb 20 2019
Pushed new branch T23104-deprecateQmitkPlotWidget.
we should add an extra view mode MaskXStatistic to avoid mixture of different things
Pushed new branch T26012-DefineTestCases.
Pushed new branch T26013-DefineGroundTruth.