[med-svn] [irtk] tag v1.0 created (now 3e7817e)
Ghislain Vaillant
ghisvail-guest at moszumanska.debian.org
Tue Mar 17 14:24:13 UTC 2015
This is an automated email from the git hooks/post-receive script.
ghisvail-guest pushed a change to tag v1.0
in repository irtk.
at 3e7817e (commit)
This tag includes the following new commits:
new 07b8c43 Created top-level directory
new d52be2d Iniitial import
new 18d6bfe Fixed bug when compiling without VTK support
new 11e3b81 Bug fix in setting the t value for the region of interest.
new f9a5fb0 Ensured weights used with global matrices. Other smaller changes.
new 5d1b102 Added Frechet weighted mean function.
new 00f7b5c Added Frechet weighted mean function.
new 5d34997 Update for better support of different image types (char, int, short, float, double)
new 837e9ab Fixed some warnings
new b12e7d8 Fixed bug in name of header file
new a3be99a Fixed bug in name of header file
new ca0208e Undo previous change
new a8f2e0e Several updates for cross-platform compatibility
new e447a52 fixed a bug
new 3de863c Updated input parameters
new 44cf718
new 9e82dec Added link to Microsoft Visual C++ 2005 Redistributable Package (x86)
new 8d0d3f2 Implemented B-spline fitting with double knots at the boundaries
new 7093c25 Constructor for control point spacing
new 1d35600 Matching mean and variance
new 4661fd3 Minor tweak
new 3a66bb3 Implemented double knots at the boundaries for B-spline fitting
new 7c141fc Matching mean and variance
new 4d74ff4 fixed bug
new 2bbdef9 removed irtkEMAffineRegistrationBiasCorrection.cc
new ac9098b Changed extensions to nifti
new b4685fb Use zero padding value when calculating histogram.
new 7453d41 Implemented bounding box for bias correction.
new f725573 Implemented bounding box for bias correction.
new 0d73223 Removed svd.
new 07a11ed Transformations were being read in twice.
new ff07a58 --add the anisotropic registration algorithm-
new 32f7a93 Fixed persistent problems with slope and rescale in NIFTI images
new c08fb8f corrected bug in irtkImageArttributes::Operator ==
new 452a986 changed options
new 23cf65b Fixed bug in reading lookup table
new 9731bfc Print parameters on screen
new b3ecacb Print parameters on screen
new 93f741b
new 7945df0 added enlarge_image
new ef95f09 updated template bias correction
new 77088c1 fixed padding value
new 3f96a6e Fixed bug for intensity display of sequences
new d728d11 Fixed bug in GetRegion
new 9a29a4b Added support for type INT
new b02e4ef Added support for type INT
new b9f3b71 Added support for MS VS 2005
new 8e74891 Changed to read images as floats instead of ints.
new 24389ba Added image to image filter for irtkImage and irtkRegionFilter
new 41b6620 Now preserves input image type
new 2c6120e Made the data type of output image match that of input.
new 3a84028 Called Run method on Region filter.
new 100f83a Minor bug fix on bounds checking.
new d3978fb Minor bug fix.
new e1bc3b4 Fixed bug
new 1c3dbe4 Fixed bug when reading fluid dofs
new c348cf0 Fixed bug when writing GIPL images
new 1b335a6 Fixed bug when writing ANALYZE images
new 554dd5e Added some command line options: -diff -tcontour -scontour -seg -lut -object_lut
new cbab49e Added routine to draw multiple objects. Thanks to JAS.
new 103acb2 Handle multiple vtkPointSet objects and a colour lookup table for drawing them. Thanks to JAS.
new 83aef64 Added functions for convolution with 1st & 2nd derivative of Gaussians
new 90d6910 Fixed bug when using target option
new 381df57 added moving source when saving movie
new a2fea66 Fixed problems with vtkPolyData display
new da0060b Fixed bug in z loops in ComputeCoefficients routine due to this->_z not being set in Initialize routine.
new 0c29483 Fixed some bugs
new b682f72 Fixed bug in initialization of coefficients image.
new 44a297b Fixed bug in screen dump code
new 8f532d3 Fixed problems with reading unsigned short images
new 779bb9d
new 62b3a99 Added support for more voxel types (unsigned short/int)
new 501bd5d Added inverse consistency as an error metric
new aa9181c Bug fixes
new 59909e2 Normals are recomputed after transformation
new 4e8bb61
new f5d4a92 Added code for model to image registration
new 93dddaa Added code for model to image registration
new 05dda61 Added code for model to image registration
new 0dad9c6 Bug fix
new c27e31b Bug fixes
new 116bacf Added code for model to image registration
new 494b747
new 5e9f609 --New advances in the diffeomorphic registration algorithm--
new c1d9561 Bug fix
new c28b550 Rewritten demons registration
new 3e41e14 Changed image format for debugging
new 4929b58 --Add a package to perform Fast Fourier Transforms (FFT) of images and the convolution of images in Fourier spaces--
new 3049909 --Small changes in the large deformations registration algorithm--
new 5bc9fc6 Fixed bugs
new 856933f Bug fixes
new e213365 Bug fixes
new 3d146c0 Bug fixes
new 57f2267 Added option for registration in plane
new 4a76fef cardiac, in case things went wrong. current result is encouraging
new 4d440b2 save recent development in the cardiac branch
new 8fc9852 delete the "branch"
new 829f31b Changed order of linking libs
new 0ec7033 Added options to fixed time info in header
new 367dca8 Minor bug fix
new dd51537 Added image filter support for other voxel types
new a29bfc1 Added support for computation of Jacobian
new d9b5433 Created branch for mcn06's GPU accelerated project
new 68dfeb6 removed bad branch
new 7ef35cf Fixed several bugs
new 333d56a Added support for more image types
new 1fb959b Fixed bug when reading NIFTI images with slope and intercept
new 890ab61 --New version of the large deformations algorithm and the fft transormations--
new 874041f Fixed bug in adding FFD to existing multi-level FFD
new 7689aa8
new feaf80e
new 8c5f490
new db24aa0 Added functions to compute displacement in base class
new 4c95166 Modified version to take multiple dofs, real valued images, a mask and to set default bin sizes.
new bf923fa Initial import.
new 025c458 Added revision info at top of file.
new 1586bb2 Added option to control whether FFDs are added or not
new c5adfbc Added option to control whether FFDs are added or not
new 1f0fb5b
new 051435e
new c341555 Ensure types remain correct when building on x86_64 Linux.
new 6421a1c Included defmap.cc
new 29bcc26 Bug fix by Sebastian Weingaertner
new 5062894 Included GPU files and modifications
new fda9fd0 Reverted accidental trunk commit
new 657c375 Imported sources
new 5c14df8 Made tmp member protected
new 58c1ba9 Added check to avoid reading images with dt = 0
new 4890dc0 Fixed bug when checking if time is within range
new 2eb89fc added function gmm-pc-par and impremented GMM fitting with more options such as uniform prior and equal variances
new 6ce6aa5
new 5d8a87e corrected bug in 'InitialiseAtlas()'
new 8cea216
new 3af967c
new 80799ab
new 8d4c294 Fixed bug when switching between transformations
new eb14d70 --Completely new version of largedeformations--
new fc7c034 implemented functionality for neonatal segmentation
new ae823d4 support for lcc
new 7022788 Removed PutAttributes member
new 75bbccd double irtkMeanShift::findGMvar() { double k; int mean = ValueToBin(_gm); double sigma;
new 4da3844 Fixed warnings
new c67b22b Fixed bugs of Multi Image Registration due to different orientation and origin
new 65cfc00 Fixed bugs of Multi Image Registration due to different orientation and origin.
new bca9b15 Corrected bug with _raster orientation for manual segmentation in sagital view
new d20ec25 change vector to ::vector to distinguish from std::vector so this file is compatible with opencv
new d870240 Code for composing transformations improved
new fc08873 Removed warnings
new f03ff0a Cosmetic changes
new 80a0860 Bug fix
new abd9072 Updated demons code
new 399cc0a bugfixed pv segmentation estimation for neonatal segmentation
new 1ced6e8 Generalized to different voxel types
new 7c9d7e0 Added function to save 2D histogram as image
new 7d6d0f1 --new version of largedeformations--
new aa43682 Added several new options
new efed6f6 Added combineLabels.cc
new 670fa23 Fixed compilation problem on ubuntu
new 01bf8f4 Fixed compilation problem on ubuntu
new b7266be Fixed bug in PutMinMax
new 04bcebd Fixed some compiler warnings
new 3596006 Implemented speedup for gradient computation
new 759bca9 Implemented templated version of histogram classes. Bins can be incremented by non-integral values
new 089aae2 Updated to reflect changes in histogram classes
new e056d39 Changes in histogram classes
new 7072678 Updated to reflect changes in histogram classes
new f1a069f Updated to reflect changes in histogram classes
new 0dcf9b1 Updated to reflect changes in histogram classes
new 649a47b Updated to reflect changes in histogram classes
new e070a5e Added template instantiation for irtkGenericImage<double>
new b5855e8 Updated to reflect changes in histogram classes
new 20b3dc2 Updated to reflect changes in histogram classes
new 50dd5ee add t dimension
new 1377fd8 added t and dt to file to gipl and gipl to file.
new 7e11af4 dicemetric
new 1fe3ad6 add some options for distance map
new edb0a17 fixed linux this-> problem
new 8d751cd Added line width for isolines/deformation grid
new 98d7dfc Fixed accidental submit
new f60fb87 surfacedistance
new f01d6e9 check and update _dt and _t after load to tolerate old gipl format where _dt and _t is zero. in case _dt and _t is equal or less then zero they are set to 1.
new 813e7c7 Added method for computing Jacobian with respect to DOFs
new 335addf Added method for computing Jacobian with respect to DOFs
new 963557c Added method for computing Jacobian with respect to DOFs
new 8810d0c Added a new set of registration classes in packages/registration2 and a new set of registration tools (areg2, nreg2 & rreg2) which use these new classes
new f2435bb
new 3b923c3 opencv
new 19c38ef windows 32 time.h
new 0a32a99 histogram bin error!
new 6a9d586 Fixed segmentation fault
new 69eb9d9 Fixed spelling error in file name of include file
new 758c2c4 Fixed bug due to t axis
new 9b472bd Imported source
new fcefe66
new 08802c4 Fixed bugs when compiling without -DNO_BOUNDS flag
new f9fabd8 Fixed memory bug in NMI gradient calculation
new 1d0aaaa Allow user to calculate their own weights (i.e. can override default use of Gaussian function).
new c4e8319 Fixed compiler warning
new 38060a4 Fixed warnings under linux
new d961875 Fixed warnings under linux (2nd try)
new 4b03840
new 4362bbf Added weighted averaging for overlap computation
new 8764226 Small bug fixes
new 18c5e7f Small bug fixes
new 3e3b74f Fixed some warnings
new 4317c34 Fixed some warnings
new 1cc9d76 Small bug fixes and improvements
new 0eb17a4 General update to improve C++ compatibility: Bool -> bool, True -> true, False -> true
new 10dbbbe Fixed even more compiler warnings
new d397999 Added labelStats executable.
new 697e02e Changed some ints to bools.
new 844c6bd Fixed bug when reading parameter file
new f42e386 fixed some memory leak problem due to multiple local and global lookup table
new 897e4bd Making a call to Bending energy function available.
new 1f31001 Allow a map of squared distances.
new 6f15a52 Fixed warning about unused parameters
new 09d0de8 shape based interpolation src
new 177c191 shape based interpolation include
new 3a8caa7 shape based interpolation basic application
new 58956b0 Fixed some warnings
new c94992d
new 213aa65 update some functions
new 2c58e3f update some functions
new 1429b36 fixed a bug. world to image wrong place...all mess up previously
new 1c7e189 ...>.<
new 133e9c5 Ensure correct interpolator is used in 2D case.
new 6d91ccf enable dof2image in terms of image coordinate
new 1aa6b04 fixed splitvolume and convert origin bug...
new 1d519e0 tricked the world coordinate to the image coordinate by adding option -image
new f05faad Guard against invalid option.
new 6bed4cf fixed -image and -center at the same time plus -image extended to areg.cc
new 4c296dd Commented out line that caused linking error.
new 7f3e3a9 fixed link error
new 7ebd6c7 fixed the msample bug in linux
new bd68ebb fixed the msample bug under windows
new 0868d8b msample bug
new a589944 move headertool
new 4b863bb move headertool.cc
new 470081f move headertool.cc
new 500e06d transform header instead of image.
new d3a0dc6 Added an arrow to deformation vector. refined the tag line drawer.
new 8518d19 --New version of the LDDMM algorithm, geodesic shooting algorithm and some tools--
new 2520e74 closestpoint in irtkpointset.
new 566c4c7 volume preservation penalty fixed
new d1fd596 fixed the mat2dof bug.
new c160b90 combine_similarity for multiple image registration
new cb00cb3 try
new 81dd8ce try again
new f373131 what about this...
new 209694f util move down!!!
new 125eb1c seems to be working now...yeah
new 128a4c0 image format
new fcbf781 rician and enlarge
new 28c8829 added option cardiac default off.
new 45a8424 Wenzhe's update package, if you found any bug please don't hesitate to contact ws207 at doc.ic.ac.uk 1 graphcut + multiple component em segmentation (segmentaion++,graphcut.cc,multiimage_graphcut.cc) 2 shape based interpolation (irtkShapeBasedInterpolationFunction,resample.cc,transformation.cc -sbased) 3 combined cardiac motion tracking (irtkCardiac3DImageFreeFormRegistration,motiontrackcardiac.cc) 4 rician distribution (irtkRician) 5 opencv image to irtk image and vice [...]
new 51d5e33 chinese?
new 45fe40c
new cded723
new c09003f cardiacdof2vtk.cc
new 1dc6afd Fixed compilation problem on Macs
new 02fc0d1 Fixed some warnings
new e6a9e98 Fixed some warnings
new 82b8459 fixed some warning and bug
new e5a5ba9 fixed some linux compile error and warnings
new 817b9ac final warnings.
new 70fe52c segment no effect >.<
new 2d325c5 added strain and fixed some warnings.
new cc34871 Version without dos carriage returns.
new 585089f Version without dos returns.
new 59b1770 Version without dos line endings. Causes difficulties with svn.
new 8cb00b1 mcubes added number of iterations of the smooth filter temporalalign fixed a bug cardiacdof2vtk.cc modified radial direction to R = N - (N dot L)L
new c977f6a combined surfacedistance with sevaluation ... similar applications
new 21fe6c8 properties
new bcbe95a properties
new c3980b9 fixed numerical volume preservation
new 830fab0 Allow executables to access data type of image data being read.
new 1ed66b9 volume preservation
new 627378d Added Revision number to usage message.
new 46eb97b Reversed previous change. Too messy.
new 548d32d Added a PrintVersion function to irtkCommon.
new 5de2e6d Added irtkVoxel.cc.
new bccf34c Added function to 'Clear' an image - i.e. deallocate data and set dimensions to zero.
new f251d9b Added option to make the output data type match the data type of the input - regardless of the data type of the target if it is specified.
new e51d9ab changed -close
new 0bcff95 set properties so header can be automatically generated.
new de21ddf header
new 7f894cc Re-wrote usage and other cosmetic changes.
new eda5105 1, added analytical computation of volume preservation penalty in registration2 2, added multi-label graphcut image segmentation, optimized memory usage by sacrificing some accuracy(0-255 instead of real number), tweeted the energy function a little 3, improved some cardiac toolkit's code
new a87bfb6 fixed padding but (3d->4d) multiple image motion track (added external weight for Xiahai) added svn property for graphcut removed bin for multiple image since every image use one metric for its self now.
new b8a3fb2 fixed non-vtk in some files
new da62352
new 1dfa6ed
new 9ced72b Trying to add CDash
new 8e65fd9 fixed some vtk problem
new 3b5629a fixed some warnings
new bc89428 50 warnings a time
new 100de62 const char* warnings
new e8ece56 labelcount array out of boundary
new 31ebdd8 supress string to char* warning
new f08d9dc line? array out of boundary.
new f386e00 fixed some warning (unused variables and array boundary) added ffddecompose.cc;autocontrast.cc;cardiac/applications/cardiaccurvature.cc fixed one dynamic memory overflow error. in cardiac/applications/motiontrackcardiac.cc
new 74e900a some extra warnings
new ba4ce15 warnings graphcut.cc: optimized the help motiontrackcardiac.cc: optimized the LandMarkPenalty
new a31fbd9 vtkfunction added some vtk functions ffddecompose.cc multi resolution solution motiontrack optimized LandMarkPenalty weight by using point to cell distance instead of point to point distance
new bcef4d7 fixed some vtk linux error
new f383820 ffddecompose multi level decompose more accurate!
new 88d31c3 Cosmetic.
new 6c0f5c8 Added helper class for iterating over the neighbourhood of a voxel.
new 07e58d4 Allowed different voxel neighbour connectivity to be chosen when eroding or dilating an image.
new 0260f54 decimate bug in vtk makes landmark penalty messed up!
new ffef453 hate vtk! delu2d could be wrong too!!
new fb7bc57 cardaic3Dimage long axis weighting problem
new 5550188 Fixed some compilation problems with TBB
new b67e174 fixed padding in multi image registration/include/irtkMultipleImageFreeFormRegistration.h
new 452407a changed distance measure to symmetric. might be slower.
new 2fe1388 quick fix
new 6f763d7 what happened?
new a1575d2 Extra helper function to initialize neighbourhood offsets using image dimensions.
new 516ce48 affd padding with landmarks
new 9d8d8f0 i -> l
new 869890a 1.0 / _SpeedupFactor for landmarks
new cef7879 lattice to image
new 7da4112
new 8a749c7 has_vtk in irtkCardiac3DImageFreeFormRegistration.cc
new 2e2ff65 minor modification for parameters
new 207b379 finalized curvature finalized motiontrack with landmarks
new 802b9f1 added cardiac surface compare to compare directly on surface between cmr1 cmr2
new 4a13bf5 line end
new e79faa6 finalized motiontrackcardiac!
new 9376dc5 line break!!!
new e25df69 removed a useless option fixed a bug
new f4dde3f added warning for image with 1 frame! modified the usage to be more clear
new 4528bbc cardiac curvature changed to radial curvature
new a4fea46 fixed the transformation display
new 7a06e22 Minor bug fixes
new 4b1badb Fixed bug in volume preservation code for 2D images
new d03c2b2 fixed the transformation display with update2
new 37094f9 output with end systolic frame number on line 2
new b4e17e7 fixed \\ problem
new 69b9508 changed easier to use
new 68b5b72 Changed irtkGradientImageFilter so it gracefully handles 2D images
new 987f812 Added new functions for computation of bending energy and its gradient
new 54e2292 modified some helps
new 924084b Minor changes
new 579ae27 Fixed a bug when returning the minimum value for voxels of type unsigned char
new 343978a Do not attempt to build irtkMultipleImage*Registration classes if VTK is not defined.
new d7f83ef Corrected previous silly error (VTK -> HAS_VTK).
new 070ad09 tried to optimize the registration
new a514425 optimized the loop
new 0a41d0d Find the local spatial Jacobian for a 4D B-spline FFD.
new 48eb4e5 Added ffdcompose tool
new a99954c Extended resample program to more voxel types
new 6eb6f8c Added tool ffdinvert Added warning for tool ffdcompose Changed irtkBSplineFreeFormTransformation so that 2D and 3D FFDs are properly supported rather than just 3D
new 7149555 Minor changes in the way the type of transformation is displayed in the transformation browser
new 0595206 adjust weight
new f583651 control point can be out side the segmentation and cause segmentation error.
new c1fa8e2 no more invert in 2D
new db47962 Make sure parameters are guessed before reading any parameter file.
new 1572ec9 Major update. Several bug fixes and improvements.
new 2cc8224 Bug fix from Stefan
new c5f0a94 to make it work with new registration2, details will be corrected later. It is not a working version now.
new 57923f9 fixed a bug with target padding
new c7f8efd to make it work in all systems.
new 1b7e4d2 try to avoid using customized opencv.
new 9740b92 for general opencv
new 7353bee removed unecessary header in overlap.cc
new 73667d0 Fixed major bug
new 1529a1e simplify the code of LandMarkPenalty
new dd3574e updated VolumePreservationPenalty for registration2
new 85e50d4 removed some commented old code.
new a982042 linux line
new e76683b use irtklocator instead of vtk locator.
new 1f52951 optimized LandMarkPenalty in multiple image registration.
new e91d568 for optimized LandMarkPenalty in multiple image registration.
new 7e00aeb I hate different coding!
new 540962c coding...
new 3071d78 coding!!!
new bdb352b Added check if control point spacing is too large
new e62e668 Renamed member function VolumePreservationPenalty to VolumePreservationPenaltyGradient
new 3d5f77c Renamed member function VolumePreservationPenalty to VolumePreservationPenaltyGradient
new fac4643 Fixed some bugs in the rigid and affine registration
new f3c715a Added rreg2, areg2, nreg2 to the condor list.
new 94539d0 Allow an option to match the source data type in the output (when a target is specified that might have a different data type).
new 1d56ad4 Make axis labels stay off when -cursor option used.
new 0bd6b06 Initialise rview intensity lookup table to contain 10000 entries.
new 2725d8c a bug when don't have landmarks but LandMarkPenalty is set.
new a7151cb one can now read any 2D grey scale image with OpenCV installed!
new 1ef5646 line...
new a6535c8 Improved default selection of parameters
new f689e4b Guess this makes more sense, number of iterations equals to number of times the gradient is evaluated.
new a260dbd Changed the implementation for how a transformation is applied to the header. Also restricted to rigid transformations and affine transformations with no shearing.
new b836a88 Re-set keywords properties on file which seemed to be broken.
new f748fc7 discontinueregistration
new ec146dc ApproximateAsNew with no residual from previous transformation
new 16f70de Fixed some warning messages
new cfddbe0 changed class name from discontinue to discontinuous
new 017e0d2 fixed some warnings.
new 6e08f0a Enable multiple reflections with more types (e.g. flip x/y). Allow arbitrary image data type.
new 9654ca5 Allow arbitrary data types and match input and output data type.
new c63b12f Tidied up usage message.
new 66c7aa9 add property for combineimages.cc headertool.cc reflect.cc and convert.cc are modified so options are self contained.
new 44f5af7 headertool.cc reflect.cc and convert.cc are modified so options are self contained.
new 4eb2b0b Added comment only.
new 67c829a General tidying.
new 3d75b63 adaptive for both lambdas
new fd3efbb Rearranging the code and tidying up. Should be no change in functionality.
new 01a0059 Re-format usage message to fit into 80 columns.
new 4858b03 Changed FlipXY, FlipXZ, etc. to take an obligatory parameter which indicates whether the user wants to modify the location of the origin.
new 558ef9c Clarified usage and added parameter to Flip calls.
new 7cb170f Allow option to swap axis vectors in pairs.
new 219ef33 Enable multiple spatial reflections with more types (e.g. flip x/y). Reflections processed in a sequence in order given.
new 8a65499 separate -swapzt from convert and build a new application to contain specific image conversion problem solution
new bfde542 Fixed some small bug for very large images and also change GuessResolution to return largest not smallest voxel size as default
new ca2767a Added option to use seperate mask image for padding (areg2, rreg2)
new fe04fa3 Option to set output data type as different from input data type.
new 5a8a62d fixed a invalid pointer for caused by speadup factor in 2D FFD registration
new 2bc9fe8 fixed a crush bug when transformation is loaded, grid draw true and source image t != 0.
new 80feabf rview bug frame changed, load new image->crash! done crash when source and target image doesn't have the same frame. done
new 629b517 <0 cases...
new 97fad9c Added #ifdef flag for printing timing information
new f28ec54 Added #ifdef flag for printing timing information (added one more)
new 32a5f55 Added City Block Distance transform class.
new d1af2c0 Added header and keywords.
new 1721aec Added option to use City block distance for distance map.
new b9802c8 Changed default voxel sizes used for 2D registration
new f7287f9 Another bug fix for the default parameters
new da49f7e Fixed bug for 2D time sequences
new 27673d2 for most cases, without !=0 speed up 20%.
new 4b5890b Added support for other interpolation types
new fd09c4e Added support for more image types
new 397e8d6 Added two executables : ems combineLabels.
new f092893 Improved performance of free-form deformation classes
new 4cc77c9 Improved performance of free-form deformation classes
new c575585 Added ems and combineLabels
new fcb1d30 Fixed reading corrupted hdr images
new 0d536cb edgedetect.cc fixed a bug with sigma = 0 splitvolume.cc output in .gz format to save space
new c4c557b Changed the way irtkFluidFreeTransformation is evaluated
new 6aa3689
new 1ed7176 for 2D transformation z displacement should = 0
new 077cc76 modified a comment
new 84c5e09 Changed the way fluid transformations are evaluated
new 3023e55 Added some more options
new c545a95 Removed unnecessary print out
new 82d5e2c Added access functions for control point spacing
new f98b864 Added friend definition for irtkImageFreeFormRegistration2
new 21d8fd9 added fast weighted median selector
new 93c9062 added trace of matrix
new 457c0ae remvoed some comments added property
new 4ee3e87 eigsrt from large to small and from small to large(for spectral clustering)
new 7f2d874 modified some details about the median filter
new f0411f1 fixed a windows related compiling issue
new 71a60f3 Fixed bug to allow compilation on Mac OS.
new ca80d71 fixed a small bug caused by transfromed source around edges, the padding pattern (*ptr2target >= 0 and *ptr2transformedsource >= 0) is not consistent and similarity gradient might get residual from previous iteration.
new 24a20bb histogram details 1 allow float 2 the position of labels
new 3132fed Implemented several speedups
new 7c581a3 Modified Print() to direct all output to cout.
new 0903375 Added functions GuessParameterSliceToVolume() and GuessParameterThickSlice() which are used in class irtkReconstruction
new 149612f Added functions GuessParameterSliceToVolume() and GuessParameterThickSlice() which are used in class irtkReconstruction
new 8d8056c Object to perform reconstruction of volumetric MRI from thick slices.
new 2761782 Object to perform reconstruction of volumetric MRI from thick slices.
new d9d6fdf Added irtkReconstruction.cc
new 3bac60d Application to perform reconstruction of volumetric MRI from thick slices.
new 771e1e1 Added reconstruction.cc
new dbe8c03 added method "SwapImages"
new ea5a357 added "SwapImages"
new 31ebccb same as SSD fixed a small bug caused by transfromed source around edges, the padding pattern (*ptr2target >= 0 and *ptr2transformedsource >= 0) is not consistent and similarity gradient might get residual from previous iteration.
new f6c71c0 added the egalitarian preconditioning and fixed a bug with non-linear interpolation.
new 98c2d99 added some option to test normalization and smooth penalty
new a8c5e1f added header
new b792fd9 svn property
new 976044c Fixed bug that meant target data type was not being set.
new 57c98e4 imagehistogram fixed a bug due to data type nonlocalmedianfilter requires buffer.
new c2f76af optimized dicemetric for multiple labels
new 78f27bf Modified the Fluid and Multi-level free form transformation classes so that ffdcompose can produce estimates for a composed transformation which retain correspondence between the first and last images in the sequence. I.e. if T1 is the transformation for images A -> B and T2 is for images B -> C, then ffdcompose produces T = T2 o T1 with correct correspondence for A -> C (and affine transformations are taken into account).
new 0211b31 fixed a interpolation error in areg2 for 2D cases.
new 1e52ac3 Allow padded image data type to match input data type - previously only short integers were output.
new e47346e Whitespace fixing.
new 7a1cc33 Changed some ints to longs to deal with large files
new 959605e Allow line thickness to be set at command line.
new 27c8f77 typo
new 795948e Make the default behaviour of dmap to fill the entire image with the calculated map. If the user wants some slices to be cleared, they need to give a flag on the command line.
new 0a14a57 added functionality
new 72bd3f6 added functionality
new 8d25f6f added functionality
new 5c57aee added functionality
new bd4ed63 added functionality
new 9ed2b4b added function GetProbMapMasked to get a masked probability map
new c69b559 added function GetProbMapMasked to get a masked probability map
new eaeb7a2 added get and set functions for the time component of a 4D atlas
new a452492 uint to int for windows
new b0154b9 Subdivide not initialized
new 87bfa99 filter with padding
new f0b2d8d 1 added svn info for most of the applications and sources 2 removed discontinous registration replace by sparse nreg 3 added sourcepadding for nreg2 to avoid padding related bugs
new 68c4e1d vtk
new 286a301 Rename files to change prefix from 'itk' to 'irtk'
new c31b641 Got SWIG wrappers compiling again. They had got out of date as the rest of the project changed. Most things work but some bits still to do, see test.py.
new caeb291 Need to ensure that function definitions in abstract base class match those in derived class exactly, otherwise SWIG wrappers are not generated properly.
new 808f745 Flags -fPIC needed for swig python wrapper under linux.
new 4b4baee Commented out checking of intent code
new 3f34f12
new 8b0a98f
new 9e820a3
new f6d645d
new c8a35b2
new 61f4622
new 4a679db
new fcd8595 Added irtkManifoldLearning and irtkSpectralClustering (required for LEAP executables)
new 587389f Added executables required for LEAP package
new 2fc8db2
new ad41fe4 Added LEAP executables
new e7ec6c0
new d6407fd
new e3c6da4
new b58ecf8 bug fixes
new 438f9fe Added option to normalize with a piecewise linear function as proposed by Nyul et al.
new fbc1896
new 2ed9810
new 0fdcc5c
new 292d7ae removed bug floor(integer)
new 0839cf3 some change to sparse_nreg.cc
new 585a9ae unstructured grid evaluation
new 3febada optimized default parameter for medical images
new a0acc2d check if has previous transformation before store it to image
new 2931849 Removed LEAProbabilistic_atlas.cc
new 1c78078 Added Normalize Nyul
new a089ef6 Removed Normalize Nyul
new 466f4cf Removed Normalize Nyul
new 713ada2
new dbefa4c sparse free form registration with previous deformation
new 2f0dcd8 fixed problem from tom's case
new 4ec70f8 Clean up
new a67f6de Added graph-cuts based segmentation for subcortical brain structures
new f976332 120816
new 926eea5
new 474c020 Avoiding run-time initialization of arrays
new eafc633
new 4afe14e Fixed warnings and problems about overloaded virtual functions
new a22cd7b Added option to perform initial rigid registration before affine registration
new 3589340 -added optional use of SSD instead of NMI
new 5106229
new b842dac some fix of windows related questions.
new 2cbf8b4 added AddWeightedSample for cardiac motion tracking
new 82919d0
new cd34b50 fix the no gradient problem
new fa477da optimized the constraint for medical images
new ab87bac
new 746d90f Fixed bug in option -isotropic
new a60b824 Can set finest level for sparse ffd
new cb9d4c0 Try to fix the not so much level bug.
new c7a1c93 fix the error with mask!
new c729cab added escaling at the end and simulation of stacks
new 1187646 added escaling at the end and simulation of stacks
new 5436391 added escaling at the end and simulation of stacks
new e83d239 Fixed compilation problem on Ubuntu 12
new aa7ff03 blur.cc output datatype when found error OpenCVToImage creates header accordingly motiontrack2 with sparsenreg irtkImageFreeFormRegistration2 fixed spacing with non isotropic images
new d0fa316 line width has an effect on object
new 89d4a82 SFFD now favors group lasso
new fcdc9cc Changes to transformation header files in main classes need to be reflected in Python interface files.
new e712083 cardiac phase detection!
new dfe9627 added properties phasedetection with blur
new ef2d265 default size
new 14d9c6c a bug
new e9cf264 cardiac volume count
new 6a56cb0 fixed a bug with wrong pointer
new fcac057 fixed cardiacphasedetection, no more blurred image
new a66b58c Fix include statement of FLTK header such that it compiles with FLTK 1.3 without requiring backwards compatible link creation.
new f5fa40a Add missing explicit template instantiation for equality operator of GenericImage.
new b051cfc Remove fluid application from build configuration as it conflicts with FLTK's fluid target and is anyway deprecated.
new ede32da Fix several issues with build of IRTK with FLTK 1.3.
new 262b9c1 Improve finding of OpenCV libraries.
new 0fd94ff Make IRTKInclude.cmake work with CMake 2.6 again.
new f85043e Fix/Enhance CMake modules to find FLTK or OpenCV, respectively.
new dbcd4a0 When using landmarks to specify region, allow them to have a z-coordinate. Check order of landmarks and swap if needed.
new cc9c566 Add missing installation rule for dof2image.
new 54bfaea some options were not functioning
new cc7da4e padding not working properly when set > 0
new 3c64f15 padding with sparse FFD
new 0795b90 int to double when multiple
new 90073f3 isotropic float now.
new b3b0aa8
new 8abb4fc Document recent change of FindFLTK.cmake module.
new ec3e517 grid transformation can output grid to irtk format (vtk polydata)
new 343536a Make conversion between IRTK affine transformations and FSL (FLIRT) matrices match code in FSL.
new 2ce7309
new 34dce1e
new dbee6bf
new e9ac6ba
new 8c6782e
new f767d51 windows related bug
new 40e6085
new 629648a bug fix
new 6e97e83 Improve/fix finding of FLTK on Windows.
new f75a287 Define HAVE_ZLIB as used by the included znzlib sources if a zlib library was found.
new 41d5f5a Minor change of file header which referred to ITK (original name of library, but misleading) instead of IRTK.
new c888312 Minor edit of documentation of some methods.
new 283d7cb temporal sffd
new db81e8f
new 933b0f6 irtkImageTSFFDRegistration
new 8c70ff2 Add missing explicit template instantiation of irtkGenericImage::operator=() to irtkTemplate.h.
new 6729ba7 an obvious bug in TSFFD
new 08ee2b7 miccai2012evaluation.cc updated with time and allow 4D transformation
new 0f7d05f better guessing of parameters
new 673647c temporal solution for TSFFD
new d657907 Add OpenCV_DIR variable to CMake cache also if library not found so user can specify it.
new fe817f4 bug
new c778efd cout clearer
new 822a0d5
new 7e0ac05 computeTSFFD.cc added version control
new 501477a seperate temporal interval
new 17b6335 added bisect_dof
new 3b6ccc0
new d979cdb
new 357e159 bisect_dof moved to packages/applications
new 8e11276 added bisect_dof (moved from irtk/applications)
new 75cf44f added bisect_dof (moved from irtk/applications)
new 04626d1 removed bisect_dof, moved to packages/applications/
new b07b2f3 minor change error output
new 6bb3716 miccai2012evaluation added strain
new e75637d fixed some bug in late and irtkbep
new 23a6991 updated with scalar, can now generate some figures.
new b22f0eb faster io
new 5bf4500 vtk2txt bug
new 250c265 calculate error at imaging area
new 92f6085 1. irtkNeighbourhoodOffsets.cc and .h Add the 4-neighbourhood connectivity.
new cd798e2 Add convenience macros to packages/applications/CMakeLists.txt file to guarantee installation of executable and avoid unnecessary merge conflicts with other developers.
new 703e322 Move comment regarding fluid executable to where this executable would be added to the build in alphabetical order.
new 58b5c11 Add helper macros for addition of executable also to applications/CMakeLists.txt file.
new 4811141
new 648a1b5
new f223b75
new 7dce7df Some minor formatting changes
new fa384ce
new 02399e1 Minor: Added warning re iterations on Frechet mean function.
new 94a0649 Fixed bug in random choice when deciding on a tied voxel.
new 874cfcb not saving confidence map any more
new 874e5ed changed to normalised GD optimisation
new e7c0485 changed to normalised GD optimisation and updated robust statistics
new e89a4f5 added median filter
new 38addf1 Application atlas now always produces the same output type as is the input
new f3f748b Now handles NIFTI files of type char correctly
new a8fdceb Bug fix
new 330b0cd Application threshold now always produces the same output type as is the input
new 721f54c Added labelStats tool
new b938fe3 Added check whether images have same size
new 276eefa good
new cbd4e09 added scalar
new 8ec5ac4 fixed some bugs
new d73f4eb extended
new 5deb9fc patch match!!!
new 71b5a47 dded CMake IRTKConfig.cmake generation. This will allow external projects to link towards IRTK as common with CMake projects (find_package(IRTK ...) )
new 1c35d5d added centering of the origin in target during registration
new 953f777 added centering of the origin in target during registration
new 165b9a1 implements registration with padding
new 58a848a implements rigid registration with padding
new 40ccb44 implements rigid registration with padding
new b9383b5 implements registration with padding
new 707b28b added registration with padding
new 9222d4e implemented registration with padding and CC
new 4fa3cec implemented registration with padding and CC
new 319543a reverted to NMI for the time being
new bc3f015 corrected bug in
new 4dce7f3 corrected bug in
new ec7682d changed to registration with padding and CC, added packages
new 446069e changed to registration with padding and CC, added packages
new 562f72e changed to registration with padding and CC, added packages
new 78369c1 surface current distances
new bb7df7c Added a new similarity metric, which is the normalised gradient correlation metric. It is calculated as the inner product of the gradient orientations in two images.
new 90d059e Added a new feature (normalised gradient vector) in the gradient image filter.
new 5a1174f - the main change is to enable the compilation of IRTK with TBB: the multithreaded version of RigidRegistrationWithPadding was missing To compile with TBB, the CMakeCache.txt needs to be edited: BUILD_TBB_EXE:BOOL=ON CMAKE_CXX_FLAGS:STRING=-DTBB_DEPRECATED CMAKE_CXX_FLAGS:STRING=-DTBB_DEPRECATED
new 2606ff5 added log_prefix and -no_intensity_matching options
new 5832b7b added -no_intensity_matching options
new 84038ba added -no_intensity_matching options
new 61d613f added a first version of a GPU accelerated basic image structure. Use according flag in cmake config to activate.
new 32afa20 changed a small issue in the gpu-test
new 53cf218 bug in cmake config
new 5d5a621 Set CUDA_SDK_ROOT_DIR relative to CUDA_TOOLKIT_ROOT_DIR if not set by FindCUDA.cmake and SDK installed as part of toolkit.
new 7c4c3fd Fix previous change regarding setting of CUDA_SDK_ROOT_DIR. Make INFO message a STATUS message to not require user to press [e] with ccmake on Unix to confirm that CUDA >= 5 is used for the build.
new c82507a Add missing IRTK include path to cudaCommon++ in case of build within the IRTK.
new 912a3a5 testing cmake flags
new 73861f4 testing cmake flags
new 2bb9ac8 testing cmake flags
new 309ce2d Fix compilation of irtkCUGenericImage with GNU GCC 4.6.3.
new 3d9cbdd Fix compilation on Unix with GNU GCC 4.6.3.
new 2b4517c check for c++11 features and nvcc, please test on linux
new c711a91 added missing files
new 2aa55e4 added missing files
new ca9803c Use CMake WARNING message to display warning regarding nvcc nullptr missing.
new 3e86d86 Add include guard to nullptr.h to avoid multiple definition of nullptr.
new 1f7ad09 excluded GPL code for common build
new d1a3241 test for nullptr for nvcc < c++11
new d72e920 Make nullptr understandable for both nvcc and plain C++ compilers.
new 7026e58 Improve finding of CUDA SDK files and look for NPP also with CMake before version 2.8.9 (FindCUDA.cmake did not look for NPP before).
new 6c4655c Fix CMake syntax.
new 7832b20 Fix finding of CUDA SDK include directory.
new f6e6b77 format
new b16e16a util for forward TFFD
new 4be6d26 forward registration TFFD
new e95f3a1 applications for TFFD
new 1db95d5 updates for caridac toolkit
new 1c15e33 fix for without vtk
new 2f96085 added config for CP12
new 92b9504 added RigidRegistrationWithPadding to the common header
new 171e75b corrected boundary conditions for super-resolution, intensity matching and robust statistics. Upsampling slices in Slice to Volume registration for better performance of low resolution data.
new 2f9e761 fixed the Read function to properly take into account the slope/intercept values. dynamic_cast was previously misused as it can only cast up in the class hierarchy, not sideways. The behaviour now implemented is the following: first the image is read in the type it is on disk (reader->GetScalarType()), then it is cast to the type requested (this->GetScalarType()), then if the type requested is float or double, and if slope is not null, then we rescale the pixel value [...]
new d939811 added normalisation of the bias field
new 42f806d added normalisation of the bias field
new b7bc180 added normalisation of the bias field
new 1701225 Added a first version of a Python interface for IRTK, using cython: only rigid transformations are currently supported
new 2f02570 fixed a bug in rview to correctly load the segmentation when there is no source image
new e316454 added -label option to rview, similar to display, to activate the segmentation labels instead of the contours
new 27c60db small fixes for the cython wrapper to work on the lab machines
new c4126d2 added support for motion model equivalent to SVN
new b3036e8 added support for motion model equivalent to SVN
new 4bab88b added support for motion model equivalent to SVN
new a52edfd display functionalities and a few fixes in the cython interface
new 7e71eaa corrected bug relating to size of _alpha
new 42f6203 Fix position of documentation of irtkHomogeneousTransformation class.
new b4781eb Minor correction of documentation of methods of irtkFreeFormTransformation. Add check to irtkFreeFormTransformation::Put() if specified control point index is not negative.
new e7f393e Fix incorrect conversion of time coordinate in irtkBSplineFreeFormTransformationPeriodic::LocalTransform.
new 180acc8 Fix incorrect conversion of time coordinate in irtkBSplineFreeFormTransformation4D::LocalTransform.
new 04e9665 Add assignment operator for constant pixel value to irtkGenericImage.
new bf278ff Add Allocate and Deallocate template function for one-dimensional arrays with specified initial value for array elements.
new e5300fa Add missing include file to list in CMake configuration file.
new fed4324 Add explicit std:: namespace identifier to inline code in irtkUtil.h to silence Eclipse warnings.
new e4c696a Add VoxelType type definition to template class irtkGenericImage.
new 348168d Move irtkGeometry.h include statement from subclass headers to irtkTransformation.h to be consistent with the controversial "include everything in the base header file" paradigm used.
new 321a9d1 Minor refactoring of base irtkTransformation class.
new 92e16b4 Add static methods to irtkBaseImage to calculate coordinate transformation matrices from image attributes structure.
new fd35dcf Add new irtkBSplineFunction class and refactor irtkBSplineFreeFormTransformation4D to make use of it.
new d41bd50 Update irtkQuaternionTransformation to overload the modified methods of the base class.
new 6bff994 Commit also modifications of .cc files of irtkTransformation and irtkQuaternionTransformation (forgot before).
new 6f967b1 Add Subdivide2D(), Subdivide3D(), and Subdivide4D() to irtkBSplineFreeFormTransformation4D.
new b255600 Remove optimization compiler flag from CMAKE_CXX_FLAGS.
new 3b1ad8c Fix bug in constructor of irtkFreeFormTransformation4D.
new c5a5aa1 Add derivative value evaluation method to irtkBSplineFunction.
new 2c0657f Add FindTBB.cmake module.
new 6f43590 Raise error in ffdinfo if input transformation is not of type irtkMultiLevelFreeFormTransformation.
new e8728e4 Add -dt and -torigin options to makesequence.
new a561dd6 Fix compilation with recent Intel TBB library.
new fde5b7a Fix NIfTI I/O to also read/write temporal offset.
new 39aff13 Change makesequence to only set temporal offset and resolution explicitly if specified by the user.
new e5985f1 Remove unused third dimension from irtkImageFreeFormRegistration2::EvaluateGradient2D.
new 48df0e7 Fix constructors of irtkBSplineFreeFormTransformation4D to also work for 2D images.
new 2a84774 Add constructor to irtkBSplineFreeFormTransformation3D which takes irtkImageAttributes as argument.
new 9e06498 Separate read/write of 3D B-spline control point data from Read()/Write() interface methods.
new 3c2594c Extend irtkBSplineFunction by 3rd and further order derivatives.
new 6e9200e Remove second const keyword before nullptr class definition.
new 35638a2 Fix const-incorrectness in constructors of irtkCUGenericImage and warning regarding definition of USE_CUDA_NPP.
new 0d280da Check number of read elements in dicom2image and raw2image to ensure that fread was successful and avoid GCC warning that return value is ignored.
new 64cf751 Fix unused cpu_time_used warning in irtkImageFreeFormRegistration2::EvaluateGradient.
new 5e86c0a Fix compiler warnings (GCC).
new 7322cbd Use const pointer to access voxel data of other image in constructor of irtkCUGenericImage.
new f822d6f Add definition of TBB_DEPRECATED if BUILD_TBB_EXE is set to ON. Revert explicit use of tbb::deprecated::concurrent_queue.
new db2c4e3 Fix coordinate transformation matrices of base image again. Bug introduced while merging and pushing earlier changes upstream.
new 8d77e81 Set default mode to affine. Removed old code.
new 9603672 Tidied up and clarified code. Not sure why the source and target image options are needed but left them in.
new b217060 Add helpers for CPU parallelization of code using TBB.
new f0fa6c0 Set SVN keyword properties on new file irtkParallel.h.
new e78cbbc Changed tmp_source, tmp_target and history so that they are object attributes instead of global variables. Not all registrations are thread safe, some global variables still remain, but at least the rigid registration is safe.
new b7960a0 - slice to volume registration is now done in a parallel for over slices - added functions to interact with the reconstruction process without using reconstruction.cc
new aa6d9df removed ClearImage as it unnecessarily loops twice over the image compared to calling Initialize() then setting all pixels to 0. used the general /= operator when scaling the image as the weight is always positive, and the operator /= takes care of not dividing by zero.
new cac2686 Added a bit more translation range to sliders in transformation editor.
new 2265dc5 Added mode filter for label images.
new 8b94083 Added modefilter executable for label images.
new 03c49c3 Added keywords to modefilter.cc .
new 901949b Replace code for measuring execution time of all registration to use the new IRTK_(START|END)_TIMING macros.
new cca5d0a change to irtkRigidRegistrationWithPadding in PackageToVolume().
new 9bb27e1 add property
new 9c7f480 unit not allowed by windows
new 2337ba1 Add definition of copysign for Windows.
new 074e098 parallelised SliceToVolume registration and CoeffInit removed unnecessary flush: they impose thread synchronisation, and anyway std::endl already calls flush added -no_log option to reconstruction.cc so that we can have it all on stdout if we want
new cd708c5 re-organised the files of the cython wrapper and added more functionalities
new 1112208 making the cython wrapper easier to install
new 8095ef1 Bug fix in histogram smoothing kernel
new 8a3ab1a Bug fix in histogram smoothing kernel
new c45a27b linking to the right libpython is optional, it is useful to check that no symbol is unreferenced
new 8dddc39 Revert "Add definition of TBB_DEPRECATED if BUILD_TBB_EXE is set to ON. Revert explicit use of tbb::deprecated::concurrent_queue."
new 87cdff2 Add note to make aware of danger of using TBB_DEPRECATED.
new dce9468 Fix build by using tbb::deprecated::concurrent_queue in irtkMultiThreadedImageRigidRegistrationWithPadding.h as well.
new 1a1bfa6 added tissue segmentation for neonates/preterms using the EM algorithm with 1st and 2nd order Markov Random Fields. (first commit)
new 96c8da6 - Non rigid BSpline FFD registration using gradient-based similarity measure - New padding scheme - New binning scheme for histogram-based similarity measures - Imges for registration treated as doubles
new 6745da8 added bias correction using a mask
new c739068 added PolynomialBiasField
new b4c369a
new 169e128 No need for small constant in gradient normalisation
new 6ea5af4
new 9ed2a73 segmentation++
new c0f745d 4d transformation for these
new d131f75 stefan's float->double
new 5419ab3 fixed includes and windows compile time issues
new 47b2464 float -> double
new 0989959 using max instead of fmax now, renamed variable
new f7e2b9f Computation of min max values now ignores padding
new 3f0b3e4 Computation of min max values now ignores padding
new 31c5bb8 fix a few things
new ccf666c Make image type irtkRealPixel.
new bede084 Fixed bug in macro for condor executables.
new a189f72 win32
new 50dcc00 realimage
new 227dd0c included UINT compatibility in getOutput
new c101c96 added "unsigned int compatibility"; corrected error message
new 91b4003 added method to replace single image/probability map
new 73b22b8 Various critical bugfixes
new 1aada0f Various critical bugfixes
new 9235885 Various critical bugfixes
new 0dcfda4 added prerequisites for the IRTK Python code commented out code section broken by Stefan's changes (irtkRealImage is now of type double). In the future, such code should be in a separate module so that the core module is robust to variations in the IRTK code base.
new d8105dc vtk and ply conversion
new 02a81b0 more general vtk2txt
new 21bf4e2 fixed compilation issue without VTK
new b90d614 Replaced class irtkEMClassificationBiasCorrection to use non-parametric model for bias field regularized by Gaussian smoothing. This affects application emsbc. Derived class irtkEMClassificationTemplateBiasCorrection was removed as well as applications bc and bc2 dependent on it. This will be replaced by single free-standing application templatebiascorrect which also uses Gaussian model for biasfield.
new 68c5ad1 Change included file irtkEMClassificationTemplateBiasCorrection.h to irtkEMClassification.h
new 1bc8dcc Iterative Resampling
new 3b8540b added class irtkReconstructionb0 which includes distortion correction for EPI structural data by registration to T2 template. Also added supporting classes irtkImageFreeFormRegistrationWithPadding and irtkMultiThreadedImageFreeFormRegistrationWithPadding
new f0a0271 The bspline case of Image2dof now expects displacements images with values in world coordinates (milimeters) as it should be
new 80ea5c9 changed iterations for calculations of b0 image
new 0f781c1 fixed header for parallel_reduce: this cannot be const as we modify the object through the parallel reduce loop
new de40cd8 better
new 130369e - parallelised most of fetal motion correction code using TBB (both parallel_for and parallel_reduce) - code has been reformated aiming at fewer lines and simpler structures (eg. using pointers only when necessary, using a reference when a copy can be avoided), and simplified indentation - reconstruction.cc has the same behaviour as before, reconstructionMasking.cc has been modified to take better into account slice by slice masking - added crf code used to ref [...]
new 96658e0 BUILD_TBB_EXE is now BUILD_WITH_TBB and is available in ccmake (no longer an advanced option)
new 9d97956
new d2ddba9
new fc65a11
new 9060e90 added dummy split required to use parallel_reduce without TBB
new 04e2c53
new 4f47c02 added new slider in rview for speed of animation
new 1f6f5bf added support for writing grey PNG. The order of pixels is kept the same (before it was flipped) in order to be consistent with the output from OpenCV
new 2f3f032 unit -> int for windows
new 8c84039 unit -> int for windows
new 98b7b83 corrected bugs/changes in irtkReconstruction to preserve original functionality
new c0ba69a made the code throw irtkExcptions instead of exiting and returning 1. to preserve previous behaviour, the error message is printed BEFORE throwing the exception, that way it always appear on stderr whether the exception is caught or not. This is required so that python can catch the exceptions instead of interrupting the interactive session.
new 6fac772 replaced exit(1) by throwing an irtkException
new eb08c56 updated Python interface (cython-based) to be built via cmake + lots of code update + fetalMasking scripts (detection and segmentation)
new bacdb0f to include cython wrapping in cmake build
new 060c39c removed the old Graph Cut Optimtization code and replaced it by the latest version (3.0) available at: http://vision.csd.uwo.ca/code/
new 642efe7 needed to add a new include_directory for external/gco
new e845642 Allow 2D interpolator to work on double data.
new d345f6a added the median application to CMakeLists.txt
new 4a6d9dd fixed the "Transperancy" typo in rview with the following one-liner: perl -i -pe 's at Transperancy@Transparancy@' $(find . -name '*.[c|h]*')
new 0a51cbf fixed a bug where template maxflow() was instantiated twice
new 6e13ffb added <short> <double> template instantiation
new adb197b more template instantiation for compilation on HPC machines
new 9325c55 fixed mispelling for TRANSPARENCY...
new c1c677d fixed bugs in irtkCRF: probabilities must stay between epsilon and 1.0 to take the log, edge weights were incorrectly defined, and we must scale values (*1000) in order to work with integers instead of floating numbers.
new 6aa8769 added basic doc for the irtk python module, using Sphinx the documentation is now available online at: http://www.doc.ic.ac.uk/~kpk09/irtk/
new 5e2613f added image saturation method using quantiles, there is now a new option '-saturate' in convert.
new 7ef1b71 corrected bug with initial transformation - need to be inverted to keep original funcionality.
new 8d1e287 added a ReflecY() when saving screenshots from rview
new a4522f1
new 2f6259a Generalized to irtkRealPixel
new e8de798 Generalization: irtkRealPixel instead of double/float
new 8a01ffe set svn properties correctly
new e0aff48 set svn properties right
new c6e0baa set svn properties right. Generalization: irtkRealPixel instead of double/float
new 246ac23 Generalization: irtkRealPixel instead of double/float
new 59667dc updated functionality for distortion correction of fetal EPI in irtkReconstructionb0
new a6780c2 trac test
new 99fa79d Test removal of useless file.
new 506ffa0 the accelerated iteration using the distance stored in the "mask" caused segmentation faults. Uncommented this part for now and forwarded issue to sp2010
new c5714ca added support for char again. Was added in 763 already but got lost on the way
new f7caea3 Segfaults reported by cl6311 fixed
new 0718403 added option to use bspline reconstruction during fetal motion correction
new d2894db added option to use bspline reconstruction during fetal motion correction
new f254249 added option to use bspline reconstruction during fetal motion correction
new 1443d24 added code for biascorrection of fMRI data
new e8105f0 fixed bug
new 4fd24ba corrected bugs in irtkReconstructionb0
new 59e9262 updated MAPM related files
new dbce77b signed distance available.
new 84b0825 new files for cardiac functional analysis
new 571de13
new 1a7361a fixed a bug with no label image
new 8346bd6 improved reconstructionb0
new f9e2074 fixed more issues with reconstructionb0
new 1cc7363 corrected reconstructionb0
new 24b8bad added smoothness penalty for distortion correction to reconstructionb0
new de6e438 padding for source image
new 71b5e66 small improvements to reconstructionb0
new b656121 Set time interval correctly for 4D data.
new 980a167 Replace 'uint' with explicit 'unsigned int' as this typedef is not available to visual studio's compiler.
new b790a56 corrected include to #include "vtkKDTreePointLocator.h"
new ad14c33 - Hessian image filter incorporated - Fixed voxeltype in convert.cc - Iterative Resampling deleted
new 9e7e43b Back to resampling with padding
new 16bd364 added utility to undistort DTI using fieldmap and shim values
new 5c14db4 corrected bug
new 0bdcb00 reconstructionCuda performs regularised superresolution similar to reconstruction.cc it is implemented in CUDA and therefore a bit faster than the TBB parallized version (speedup is currently ~40 compared to single CPU). Beware that after the first build TBB might give you a warning about the compiler. It vanishes after the second build and has no effect. TODO full Integration into IRTK CMake environment -> need to build separately for now TODO optimizations
new f85a006 minor changes
new d5ef9a3 added cuda reconstruction to full build -- wil be built if BUILD_GPU_SUPPORT_WITH_CUDE is enabled
new cd85314 small changes
new 723f53b mean filter
new ebd1d0f small change
new 11364dc cardiac tools
new 6eb7791 In reconstruction.cc changed stacks trasformations to be an optional input, plus reduced output for -debug option.
new a6a092d added time
new aafbee2 updated some description, previous was not correct
new 6e1a6fc fix small bug of duplicated coarse levels, 20% speed up
new 536c95c Fixed bug produced by q0, q1, and saturate flag not being initialised
new 955724f final patchmatch segmentation
new 5144a89 forgot the file
new 7ec653c Minor bug fix to allow compilation on non-windows platforms.
new 5514b28 minor cleanup, more to come
new 7a0dece updated linear interpolation to better handle boundaries
new af3f2b7 added code to compute two quantiles from an image added Saturate as a pure virtual function of irtkBaseImage
new 57b01e5 added rescale option (constrast stretch) to rreg to overcome the limitations of irtkGreyImage (limited range of intensity values)
new b3db417 updated doxygen configuration using Andreas' configuration
new c58b5e0 I slightly updated rview's options so that you can put some more colors in your life. The following line will put a pink background, green contour for the source and red contour for the target:
new d2cbff4 checked into SVN the version of reconstructionMasking used for NeuroImage paper
new ccd6891 updated IRTK-cython in particular removed matplotlib dependency updated detection and segmentation scripts used in NeuroImage paper
new 3c54279 updated documentation of IRTK-cython
new d3ba37c Simplified AdaptiveSmoothing()
new 633b211 removed class irtkReconstructionb0
new bafaab6 removed class irtkReconstructionb0
new 1a002a8 changed -line_thickness to -line in rview and added it to the help usage.
new 4a7ff52 added irtkGaussianBlurring to the cython interface
new 76586ae final commit before moving to git.
new ada17f7 Fix (presumable) bug in irtkBSplineFreeFormTransformation3D::ApproximateAsNew. Compare to irtkBSplineFreeFormTransformation3D::Approximate.
new 0ad2ad6 Added comments to dofcombine.cc
new 6f5bc21 add comments to dofcombine.cc
new 4a4b425 Add dofaverage.cc for computing the average of affine transformations.
new 5e53f17 Issue #3 Fix build of cardiac package without VTK library.
new 391b768 Merge branch 'develop' of /home/git/repositories/as12312/irtk into develop
new 2cd20b6 #2 Set CMake CMP0022 policy to avoid warning with 2.8.12.
new 76f154a Re-write ffdcompose.cc. The old version implements the addition of two multi-level FFDs by pushing all the local FFDs together. The re-written version implements the composition of two multi-level FFDs.
new 62920e0 Comment on ffdcompose.cc
new ca06be7 Merge branch 'feature' into develop
new e88f1cb Merge branch 'develop' of https://gitlab.doc.ic.ac.uk/sk1712/irtk into develop
new 3d44674 Fix bug in ffdcompose.cc
new e42ad85 Add Google C++ Testing Framework as a submodule
new dea89dc Add unit test support through Google Test
new 166c4ae Fix permissions
new da7f92f Refactored application combineLabels to install unit tests on it
new 4d4d87a Clean C++ aspects of combineLabels
new fb918d3 Update applications/ CMakeLists.txt to add Boost support
new 82c3d74 Clear applications/combineLabels
new f9f1908 Merge branch 'removeRecipes_combineLabels.cc' into refactor
new bf919af Enforce header inclusions to break inclusion order dependency
new 7fec8d8 Remove TODO following Andrea's comment
new f6665c1 Unit test irtPointSet
new 34608be Transform packages/applications/jactest.cc into a test for geometry++/irtkAffineTransform.cc
new 2f3dc96 Merge branch 'test_irtkPointSet' into refactor
new 826eaf7 Remove unused vegas algorithm implementation
new 324ef56 Merge branch 'remove_unused_vegas_algorithm' into refactor
new 68aba79 refactored application makevolume to write unit tests
new 3040535 Wrote unit tests for makevolume
new 54772e9 Added random test
new f287281 Replaced indexx in makevolume application
new 1f67329 Fixed call to indexing function in makevolume.cc
new 88a87a2 Added allocation of index array in makevolume.cc
new e97c70b Remove redundant includes of recipes
new e76c880 Wrote tests for sort function in irtkGraphCutSegmentation_4D.cc
new c38fa0b Replaced recipes sort function in irtkGraphCutSegmentation_4D.cc
new 699e5d8 Merge branch 'removeRevipes_irtkGraphCutSegmentation_4D.cc' into refactor
new 3f0db84 Wrote tests for weightedmedian.cc
new 5396803 Replaced recipes sort2 function in weightedmedian.cc
new cbcdd60 Merge branch 'removeRecipes_weightedmedian.cc' into refactor
new e224986 Removed recipes header inclusion from weightedmedian.cc
new e67e333 Replaced recipes sort2 function from ffdedit.cc
new fce89b3 Replace recipes gasdev function in irtkGaussianNoise.cc
new c0b6fc8 Replace recipes gasdev function in irtkGaussianNoise.cc
new cf7ed4b Removed recipes gasdev function from irtkRicianNoise.cc
new a30d172 Removed recipes gasdev function from ffdadd.cc
new 1916cc2 Unit test for gaussian deviates generator
new aa9d236 Removed recipes ran2 from irtkUniformNoise.cc
new 7052c0d Replaced recipes ran2 function in irtkModeFilter.cc
new 2214f94 Unit tests for irtkRician.cc
new 5f334da Replaced recipes Bessel functions in irtkRician.cc
new 7c033c2 Merge branch 'removeRecipes_irtkRician.cc' into refactor
new 308af78 Added GSL support
new d2c56cb Replaced recipes functions in irtkMatrix.Det irtkMatrix.Adjugate and irtk Matrix.Transpose
new 4cd485d Unit tests for irtkMatrix.Det irtkMatrix.Adjugate and irtkMatrix.Invert
new 13af162 Unit tests for irtkMatrix.SVD and irtkMatrix.LeastSquaresFit
new d2c984d Platform specific include libs
new 627af9c Replace conversions to/from NR data structures with GSL data structures
new 0534f3c Unit tests for irtkMatrix.Eigenvalues
new ff19256 Update python wrapper for irtkMatrix
new 9ea9194 Replace NR functions in irtkMatrix with GSL functions
new 1b76ac0 Update irtkVector python wrapper to exclude NR
new 1897c7e Unit test for Newton Method
new b99b30e Replaced NR Newton Method with GSL
new 87e7c6e Remove redundant definition in newt2.cc
new c9997df Remove redundant inclusions from newt_test.cc
new b2d85ce irtkMatrix.Eigenvalues throw exception when matrix is not symmetric
new f1d0b8c Unit test for irtkConjugateGradientDescentOptimizer
new d66af5f Replace recipes frprmn function for minimization with GSL Polak-Ribiere algorithm
new a15b34a Unit test for irtkSteepestGradientDescentOptimizer
new 5b270b6 Remove recipes linmin function from irtkSteepestGradientDescentOptimizer and replace with GSL steepest descent routine
new e0d2dc7 Modify irtkOptimizer.cc for GSL integration
new 41cda0e Remove recipes library from IRTK project
new 8bbfdb7 Remove unused variable declarations
new 4d78915 Update applications license details
new 45b524b Fix VTK version-specific commands
new eeb9e17 Add common++ license details
new fadeb57 Including Boost library commands transferred to IRTKInclude
new b196c0c Add contrib++ license details
new 424c8cb Remove empty lines at the end of CMakeLists file
new 7eb293c Add geometry++ license details
new 81d935d Add image++ license info
new 470ff84 VTK version specific commands for irtkGenericImage
new 3afe595 Add packages license info
new c638495 Fix VTK version specific commands in packages
new 4c5c3d8 Update computeTFFD.cc
new 2d261d4 Update computeTSFFD.cc
new 72ce2a4 Add license details for packages applications
new b3bce67 Update irtkBSplineFreeFormTransformationPeriodic.cc fix bug for 2D sequence
new 55c12a3 Update license and copyright files
new ff5a622 Modified file paths for registration tests
new 20f0f57 Require Boost 1.48 as minimum version
new f8ec46b Create .travis.yml
new 3eadfd9 Update .travis.yml
new 55de541 Update .travis.yml
new 97599a6 Update .travis.yml
new 70952ad Update .travis.yml
new 733bdb1 Create README.md
new 3e7817e Delete README
The 1013 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "adds" were already present in the repository and have only
been added to this reference.
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/debian-med/irtk.git
More information about the debian-med-commit
mailing list