-
Notifications
You must be signed in to change notification settings - Fork 42
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
28 sliding regions #40
base: master
Are you sure you want to change the base?
Conversation
created reg_f3d_sli .h and .cpp files updated CMakeLists to build new class
tidied up GetDeformationField() and GetSimilarityMeasureGradient() methods
already implemented changed reg_base so that methods for allocating warped, def field, and warped gradent are called after calling method to initialise current level as this method assigns currentDistanceMap which is required to allocate the warped distance maps which are allocated in AllocateWarped method
modified reg_f3d (executable) to parse and set sliding regions options ready to test
…oth regions now done
made GetWarpedImage method work with sliding registration results code now saves a result, but it was 'blank', so still more bugs to find and fix!
…' and 'region2' to name of saved CPGs
… mask supplied corrected GetSimilarityMeasureGradient() method to work with vector field previously gap-overlap constraint term value was normalised by number of voxels, but gradient was not now distance map image is pre-normalised during initialisation so that both the gap-overlap constraint term and its gradient are normalised note - weighted distance map is divided by square root of number of voxels, as the two warped distance maps are mutliplied together this will normalise the gap-overlap constraint term by the number of voxels also added some debugging comments which are commented out but still in code
…dat/niftyreg.git into 28-sliding-regions Conflicts: niftyreg_build_version.txt
sliding region registrations now appear to be working, but sensible default gap-overlap constraint term weight still needs to be determined
Finally integrated the CI to this branch. Can you have a look? |
Have fixed the compilation error and successfully tested for comparable results on Windows and Linux (VM). CI-pipeline on gitlab now stuck with missing Mac-build. @mmodat can you check that you are happy once the Mac-build is complete? Thanks!!! |
In the meantime, some merge conflicts came up which are now solved. Can you have a look and merge, please? |
Sliding region implementation completed as presented at WBIR 2018 in Leiden, NL. References added.