diff options
| -rw-r--r-- | Readme.md | 31 | ||||
| -rwxr-xr-x[-rw-r--r--] | build/run.sh | 23 | ||||
| -rw-r--r-- | demos/SoftwareX_supp/Demo_RealData_Recon_SX.py | 18 | ||||
| -rw-r--r-- | demos/demoMatlab_3Ddenoise.m | 16 | ||||
| -rwxr-xr-x | run.sh | 26 | ||||
| -rw-r--r-- | src/Matlab/mex_compile/compileGPU_mex.m | 5 | ||||
| -rw-r--r-- | src/Matlab/mex_compile/regularisers_CPU/PatchSelect.c | 4 |
7 files changed, 53 insertions, 70 deletions
@@ -1,12 +1,11 @@ # CCPi-Regularisation Toolkit (CCPi-RGL) - - | Master | Development | Anaconda binaries | |--------|-------------|-------------------| | [](https://anvil.softeng-support.ac.uk/jenkins/job/CILsingle/job/CCPi-Regularisation-Toolkit/) | [](https://anvil.softeng-support.ac.uk/jenkins/job/CILsingle/job/CCPi-Regularisation-Toolkit-dev/) |   [ ](https://anaconda.org/ccpi/ccpi-regulariser) | -**Iterative image reconstruction (IIR) methods normally require regularisation to stabilise the convergence and make the reconstruction problem (inverse problem) more well-posed. The CCPi-RGL software provides 2D/3D and multi-channel regularisation strategies to ensure better performance of IIR methods. The regularisation modules are well-suited to use with [splitting algorithms](https://en.wikipedia.org/wiki/Augmented_Lagrangian_method#Alternating_direction_method_of_multipliers), such as, [ADMM](https://github.com/dkazanc/ADMM-tomo) and [FISTA](https://github.com/dkazanc/FISTA-tomo). Furthermore, the toolkit can be used for simpler inversion tasks, such as, image denoising, inpaiting, deconvolution etc. The core modules are written in C-OMP and CUDA languages and wrappers for Matlab and Python are provided.** +**Iterative image reconstruction (IIR) methods frequently require regularisation to ensure convergence and make inverse problem well-posed. The CCPi-RGL toolkit provides a set of 2D/3D regularisation strategies to guarantee a better performance of IIR methods (higher SNR and resolution). The regularisation modules for scalar and vectorial datasets are based on the [proximal operator](https://en.wikipedia.org/wiki/Proximal_operator) framework and can be used with [proximal splitting algorithms](https://en.wikipedia.org/wiki/Proximal_gradient_method), such as PDHG, Douglas-Rachford, ADMM, FISTA and [others](https://arxiv.org/abs/0912.3522). While the main target for CCPi-RGL is [tomographic image reconstruction](https://github.com/dkazanc/TomoRec), the toolkit can be used for image denoising and inpaiting problems. The core modules are written in C-OMP and CUDA languages and wrappers for Matlab and Python are provided.** + <div align="center"> <img src="demos/images/probl.png" height="225"><br> @@ -20,7 +19,7 @@ <img src="demos/images/TV_vs_NLTV.jpg" height="300"><br> </div> -## Prerequisites: +## Prerequisites: * [MATLAB](www.mathworks.com/products/matlab/) OR * Python (tested ver. 3.5/2.7); Cython @@ -29,7 +28,7 @@ ## Package modules: -### Single-channel (denoising): +### Single-channel (scalar): 1. Rudin-Osher-Fatemi (ROF) Total Variation (explicit PDE minimisation scheme) **2D/3D CPU/GPU** (Ref. *1*) 2. Fast-Gradient-Projection (FGP) Total Variation **2D/3D CPU/GPU** (Ref. *2*) 3. Split-Bregman (SB) Total Variation **2D/3D CPU/GPU** (Ref. *5*) @@ -39,7 +38,7 @@ 7. A joint ROF-LLT (Lysaker-Lundervold-Tai) model for higher-order regularisation **2D/3D CPU/GPU** (Ref. *10,11*) 8. Nonlocal Total Variation regularisation (GS fixed point iteration) **2D CPU/GPU** (Ref. *12*) -### Multi-channel (denoising): +### Multi-channel (vectorial): 1. Fast-Gradient-Projection (FGP) Directional Total Variation **2D/3D CPU/GPU** (Ref. *3,4,2*) 2. Total Nuclear Variation (TNV) penalty **2D+channels CPU** (Ref. *7*) @@ -68,10 +67,10 @@ build/jenkins-build.sh this will install `conda build` environment and compiles C/C++ and Python wrappers and performs basic tests for environment with python 3.6 and numpy 1.12. ### CMake -If you want to build directly using cmake, install CMake (v.>=3) to configure it. Additionally you will need a C compiler, `make` (on linux) and CUDA SDK where available. The toolkit may be used directly from C/C++ as it is compiled as a shared library (check-out the include files in `Core` for this) -1. Clone this repository to a directory, i.e. `CCPi-Regularisation-Toolkit`, -2. create a build directory. -3. Issue `cmake` to configure (or `cmake-gui`, or `ccmake`, or `cmake3`). Use additional flags to fine tune the configuration. +If you want to build directly using cmake, install CMake (v.>=3) to configure it. Additionally you will need a C compiler, `make` (on linux) and CUDA SDK where available. The toolkit may be used directly from C/C++ as it is compiled as a shared library (check-out the include files in `Core` for this) +1. Clone this repository to a directory, i.e. `CCPi-Regularisation-Toolkit`, +2. create a build directory. +3. Issue `cmake` to configure (or `cmake-gui`, or `ccmake`, or `cmake3`). Use additional flags to fine tune the configuration. Flags used during configuration @@ -119,7 +118,7 @@ conda install ccpi-regulariser -c ccpi -c conda-forge #### Python build -If passed `CONDA_BUILD=ON` the software will be installed by issuing `python setup.py install` which will install in the system python (or whichever other python it's been picked up by CMake at configuration time.) +If passed `CONDA_BUILD=ON` the software will be installed by issuing `python setup.py install` which will install in the system python (or whichever other python it's been picked up by CMake at configuration time.) If passed `CONDA_BUILD=OFF` the software will be installed in the directory pointed by `${PYTHON_DEST_DIR}` which defaults to `${CMAKE_INSTALL_PREFIX}/python`. Therefore this directory should be added to the `PYTHONPATH`. If Python is not picked by CMake you can provide the additional flag to CMake `-DPYTHON_EXECUTABLE=/path/to/python/executable`. @@ -128,12 +127,12 @@ If Python is not picked by CMake you can provide the additional flag to CMake `- Matlab wrapper will install in the `${MATLAB_DEST_DIR}` directory, which defaults to `${CMAKE_INSTALL_PREFIX}/matlab` -If Matlab is not picked by CMake, you could add `-DMatlab_ROOT_DIR=<Matlab directory>`. +If Matlab is not picked by CMake, you could add `-DMatlab_ROOT_DIR=<Matlab directory>`. #### Linux Because you've installed the modules in `<your favourite install directory>` you need to instruct Matlab to look in those directories: -```bash +```bash PATH="/path/to/mex/:$PATH" LD_LIBRARY_PATH="/path/to/library:$LD_LIBRARY_PATH" matlab ``` @@ -147,8 +146,8 @@ addpath(/path/to/library); #### Legacy Matlab installation (partly supported, please use Cmake) ``` - - cd /Wrappers/Matlab/mex_compile + + cd src/Matlab/mex_compile compileCPU_mex.m % to compile CPU modules compileGPU_mex.m % to compile GPU modules (see instructions in the file) ``` @@ -179,7 +178,7 @@ addpath(/path/to/library); 12. [Abderrahim E., Lezoray O. and Bougleux S. 2008. Nonlocal discrete regularization on weighted graphs: a framework for image and manifold processing." IEEE Trans. Image Processing 17(7), pp. 1047-1060.](https://ieeexplore.ieee.org/document/4526700) ### References to Software: -* If software is used, please refer to [11], however, the supporting publication is in progress. +* If software is used, please refer to [11], however, the supporting publication is in progress. ### Applications: diff --git a/build/run.sh b/build/run.sh index d450299..f2869e5 100644..100755 --- a/build/run.sh +++ b/build/run.sh @@ -5,20 +5,23 @@ rm -r build_proj # pip install cython mkdir build_proj cd build_proj/ -make clean +#make clean export CIL_VERSION=19.03 # install Python modules without CUDA cmake ../ -DBUILD_PYTHON_WRAPPER=ON -DBUILD_MATLAB_WRAPPER=OFF -DBUILD_CUDA=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install # install Python modules with CUDA # cmake ../ -DBUILD_PYTHON_WRAPPER=ON -DBUILD_MATLAB_WRAPPER=OFF -DBUILD_CUDA=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install +# install Matlab modules without CUDA +#cmake ../ -DBUILD_PYTHON_WRAPPER=OFF -DMatlab_ROOT_DIR=/dls_sw/apps/matlab/r2014a/ -DBUILD_MATLAB_WRAPPER=ON -DBUILD_CUDA=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install # install Matlab modules with CUDA -# cmake ../ -DBUILD_PYTHON_WRAPPER=OFF -DMatlab_ROOT_DIR=/dls_sw/apps/matlab/r2014a/ -DBUILD_MATLAB_WRAPPER=ON -DBUILD_CUDA=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install -make install -#### Python +#cmake ../ -DBUILD_PYTHON_WRAPPER=OFF -DMatlab_ROOT_DIR=/dls_sw/apps/matlab/r2014a/ -DBUILD_MATLAB_WRAPPER=ON -DBUILD_CUDA=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install +############### Python(linux)############### #cp install/lib/libcilreg.so install/python/ccpi/filters -cd install/python -export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:../lib -spyder -##### one can also run Matlab in Linux as: -#PATH="/path/to/mex/:$PATH" LD_LIBRARY_PATH="/path/to/library:$LD_LIBRARY_PATH" matlab -#PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build/install/matlab/:$PATH" LD_LIBRARY_PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build/install/lib:$LD_LIBRARY_PATH" matlab +# cd install/python +# export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:../lib +# spyder +############### Matlab(linux)############### +### export LD_PRELOAD=/home/algol/anaconda3/lib/libstdc++.so.6 # if there is libstdc error in matlab +# PATH="/path/to/mex/:$PATH" LD_LIBRARY_PATH="/path/to/library:$LD_LIBRARY_PATH" matlab +# PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build_proj/install/matlab/:$PATH" LD_LIBRARY_PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build_proj/install/lib:$LD_LIBRARY_PATH" matlab +# PATH="/home/algol/Documents/DEV/CCPi-Regularisation-Toolkit/build_proj/install/matlab/:$PATH" LD_LIBRARY_PATH="/home/algol/Documents/DEV/CCPi-Regularisation-Toolkit/build_proj/install/lib:$LD_LIBRARY_PATH" /home/algol/SOFT/MATLAB9/bin/matlab diff --git a/demos/SoftwareX_supp/Demo_RealData_Recon_SX.py b/demos/SoftwareX_supp/Demo_RealData_Recon_SX.py index ca8f1d2..5991989 100644 --- a/demos/SoftwareX_supp/Demo_RealData_Recon_SX.py +++ b/demos/SoftwareX_supp/Demo_RealData_Recon_SX.py @@ -1,15 +1,15 @@ #!/usr/bin/env python3 # -*- coding: utf-8 -*- """ -This demo scripts support the following publication: -"CCPi-Regularisation Toolkit for computed tomographic image reconstruction with +This demo scripts support the following publication: +"CCPi-Regularisation Toolkit for computed tomographic image reconstruction with proximal splitting algorithms" by Daniil Kazantsev, Edoardo Pasca, Martin J. Turner, Philip J. Withers; Software X, 2019 ____________________________________________________________________________ * Reads real tomographic data (stored at Zenodo) --- https://doi.org/10.5281/zenodo.2578893 * Reconstructs using TomoRec software -* Saves reconstructed images +* Saves reconstructed images ____________________________________________________________________________ >>>>> Dependencies: <<<<< 1. ASTRA toolbox: conda install -c astra-toolbox astra-toolbox @@ -40,7 +40,7 @@ data_norm = normaliser(dataRaw, flats, darks, log='log') del dataRaw, darks, flats intens_max = 2.3 -plt.figure() +plt.figure() plt.subplot(131) plt.imshow(data_norm[:,150,:],vmin=0, vmax=intens_max) plt.title('2D Projection (analytical)') @@ -72,7 +72,7 @@ FBPrec = RectoolsDIR.FBP(data_norm[0:100,:,det_y_crop]) sliceSel = 50 max_val = 0.003 -plt.figure() +plt.figure() plt.subplot(131) plt.imshow(FBPrec[sliceSel,:,:],vmin=0, vmax=max_val, cmap="gray") plt.title('FBP Reconstruction, axial view') @@ -108,7 +108,7 @@ RectoolsIR = RecToolsIR(DetectorsDimH = np.size(det_y_crop), # DetectorsDimH # DetectorsDimV = 100, # DetectorsDimV # detector dimension (vertical) for 3D case only AnglesVec = angles_rad, # array of angles in radians ObjSize = N_size, # a scalar to define reconstructed object dimensions - datafidelity='LS',# data fidelity, choose LS, PWLS (wip), GH (wip), Student (wip) + datafidelity='LS',# data fidelity, choose LS, PWLS, GH (wip), Students t (wip) nonnegativity='ENABLE', # enable nonnegativity constraint (set to 'ENABLE') OS_number = None, # the number of subsets, NONE/(or > 1) ~ classical / ordered subsets tolerance = 0.0, # tolerance to stop inner (regularisation) iterations earlier @@ -124,7 +124,7 @@ RecADMM_reg_sbtv = RectoolsIR.ADMM(data_norm[0:100,:,det_y_crop], sliceSel = 50 max_val = 0.003 -plt.figure() +plt.figure() plt.subplot(131) plt.imshow(RecADMM_reg_sbtv[sliceSel,:,:],vmin=0, vmax=max_val, cmap="gray") plt.title('3D ADMM-SB-TV Reconstruction, axial view') @@ -164,7 +164,7 @@ RecADMM_reg_rofllt = RectoolsIR.ADMM(data_norm[0:100,:,det_y_crop], sliceSel = 50 max_val = 0.003 -plt.figure() +plt.figure() plt.subplot(131) plt.imshow(RecADMM_reg_rofllt[sliceSel,:,:],vmin=0, vmax=max_val) plt.title('3D ADMM-ROFLLT Reconstruction, axial view') @@ -202,7 +202,7 @@ RecADMM_reg_tgv = RectoolsIR.ADMM(data_norm[0:100,:,det_y_crop], sliceSel = 50 max_val = 0.003 -plt.figure() +plt.figure() plt.subplot(131) plt.imshow(RecADMM_reg_tgv[sliceSel,:,:],vmin=0, vmax=max_val) plt.title('3D ADMM-TGV Reconstruction, axial view') diff --git a/demos/demoMatlab_3Ddenoise.m b/demos/demoMatlab_3Ddenoise.m index ec0fd88..6b21e86 100644 --- a/demos/demoMatlab_3Ddenoise.m +++ b/demos/demoMatlab_3Ddenoise.m @@ -18,9 +18,10 @@ Ideal3D(:,:,i) = Im; end vol3D(vol3D < 0) = 0; figure; imshow(vol3D(:,:,7), [0 1]); title('Noisy image'); -lambda_reg = 0.03; % regularsation parameter for all methods + %% fprintf('Denoise a volume using the ROF-TV model (CPU) \n'); +lambda_reg = 0.03; % regularsation parameter for all methods tau_rof = 0.0025; % time-marching constant iter_rof = 300; % number of ROF iterations epsil_tol = 0.0; % tolerance @@ -31,14 +32,17 @@ fprintf('%s %f \n', 'RMSE error for ROF is:', rmse_rof); figure; imshow(u_rof(:,:,7), [0 1]); title('ROF-TV denoised volume (CPU)'); %% % fprintf('Denoise a volume using the ROF-TV model (GPU) \n'); +% lambda_reg = 0.03; % regularsation parameter for all methods % tau_rof = 0.0025; % time-marching constant % iter_rof = 300; % number of ROF iterations -% tic; u_rofG = ROF_TV_GPU(single(vol3D), lambda_reg, iter_rof, tau_rof); toc; +% epsil_tol = 0.0; % tolerance +% tic; u_rofG = ROF_TV_GPU(single(vol3D), lambda_reg, iter_rof, tau_rof, epsil_tol); toc; % rmse_rofG = (RMSE(Ideal3D(:),u_rofG(:))); % fprintf('%s %f \n', 'RMSE error for ROF is:', rmse_rofG); % figure; imshow(u_rofG(:,:,7), [0 1]); title('ROF-TV denoised volume (GPU)'); %% fprintf('Denoise a volume using the FGP-TV model (CPU) \n'); +lambda_reg = 0.03; % regularsation parameter for all methods iter_fgp = 300; % number of FGP iterations epsil_tol = 0.0; % tolerance tic; [u_fgp,infovec] = FGP_TV(single(vol3D), lambda_reg, iter_fgp, epsil_tol); toc; @@ -47,9 +51,10 @@ rmse_fgp = (RMSE(Ideal3D(:),u_fgp(:))); fprintf('%s %f \n', 'RMSE error for FGP-TV is:', rmse_fgp); figure; imshow(u_fgp(:,:,7), [0 1]); title('FGP-TV denoised volume (CPU)'); %% -% fprintf('Denoise a volume using the FGP-TV model (GPU) \n'); +fprintf('Denoise a volume using the FGP-TV model (GPU) \n'); +% lambda_reg = 0.03; % regularsation parameter for all methods % iter_fgp = 300; % number of FGP iterations -% epsil_tol = 1.0e-05; % tolerance +% epsil_tol = 0.0; % tolerance % tic; u_fgpG = FGP_TV_GPU(single(vol3D), lambda_reg, iter_fgp, epsil_tol); toc; % rmse_fgpG = (RMSE(Ideal3D(:),u_fgpG(:))); % fprintf('%s %f \n', 'RMSE error for FGP-TV is:', rmse_fgpG); @@ -66,7 +71,7 @@ figure; imshow(u_sb(:,:,7), [0 1]); title('SB-TV denoised volume (CPU)'); %% % fprintf('Denoise a volume using the SB-TV model (GPU) \n'); % iter_sb = 150; % number of SB iterations -% epsil_tol = 1.0e-05; % tolerance +% epsil_tol = 0.0; % tolerance % tic; u_sbG = SB_TV_GPU(single(vol3D), lambda_reg, iter_sb, epsil_tol); toc; % rmse_sbG = (RMSE(Ideal3D(:),u_sbG(:))); % fprintf('%s %f \n', 'RMSE error for SB-TV is:', rmse_sbG); @@ -88,6 +93,7 @@ figure; imshow(u_rof_llt(:,:,7), [0 1]); title('ROF-LLT denoised volume (CPU)'); % lambda_LLT = lambda_reg*0.35; % LLT regularisation parameter % iter_LLT = 300; % iterations % tau_rof_llt = 0.0025; % time-marching constant +% epsil_tol = 0.0; % tolerance % tic; u_rof_llt_g = LLT_ROF_GPU(single(vol3D), lambda_ROF, lambda_LLT, iter_LLT, tau_rof_llt, epsil_tol); toc; % rmse_rof_llt = (RMSE(Ideal3D(:),u_rof_llt_g(:))); % fprintf('%s %f \n', 'RMSE error for ROF-LLT is:', rmse_rof_llt); @@ -1,26 +0,0 @@ -#!/bin/bash -echo "Building CCPi-regularisation Toolkit using CMake" -rm -r build_proj -# Requires Cython, install it first: -# pip install cython -mkdir build_proj -cd build_proj/ -#make clean -export CIL_VERSION=19.03 -# install Python modules without CUDA -# cmake ../ -DBUILD_PYTHON_WRAPPER=ON -DBUILD_MATLAB_WRAPPER=OFF -DBUILD_CUDA=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install -# install Python modules with CUDA -# cmake ../ -DBUILD_PYTHON_WRAPPER=ON -DBUILD_MATLAB_WRAPPER=OFF -DBUILD_CUDA=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install -# install Matlab modules without CUDA -#cmake ../ -DBUILD_PYTHON_WRAPPER=OFF -DMatlab_ROOT_DIR=/dls_sw/apps/matlab/r2014a/ -DBUILD_MATLAB_WRAPPER=ON -DBUILD_CUDA=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install -# install Matlab modules with CUDA -cmake ../ -DBUILD_PYTHON_WRAPPER=OFF -DMatlab_ROOT_DIR=/dls_sw/apps/matlab/r2014a/ -DBUILD_MATLAB_WRAPPER=ON -DBUILD_CUDA=ON -DCMAKE_BUILD_TYPE=Release -DCMAKE_INSTALL_PREFIX=./install -make install -#### Python -#cp install/lib/libcilreg.so install/python/ccpi/filters -# cd install/python -# export LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:../lib -# spyder -##### Matlab (Linux) -#PATH="/path/to/mex/:$PATH" LD_LIBRARY_PATH="/path/to/library:$LD_LIBRARY_PATH" matlab -PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build_proj/install/matlab/:$PATH" LD_LIBRARY_PATH="/home/kjy41806/Documents/SOFT/CCPi-Regularisation-Toolkit/build_proj/install/lib:$LD_LIBRARY_PATH" matlab diff --git a/src/Matlab/mex_compile/compileGPU_mex.m b/src/Matlab/mex_compile/compileGPU_mex.m index 3a7ac7c..7e15233 100644 --- a/src/Matlab/mex_compile/compileGPU_mex.m +++ b/src/Matlab/mex_compile/compileGPU_mex.m @@ -4,7 +4,7 @@ % In order to compile CUDA modules one needs to have nvcc-compiler % installed (see CUDA SDK), check it under MATLAB with !nvcc --version -% In the code bellow we provide a full explicit path to nvcc compiler +% In the code bellow we provide a full explicit path to nvcc compiler % ! paths to matlab and CUDA sdk can be different, modify accordingly ! % Tested on Ubuntu 18.04/MATLAB 2016b/cuda10.0/gcc7.3 @@ -68,7 +68,8 @@ movefile('LLT_ROF_GPU.mex*',Pathmove); delete TV_ROF_GPU_core* TV_FGP_GPU_core* TV_SB_GPU_core* dTV_FGP_GPU_core* NonlDiff_GPU_core* Diffus_4thO_GPU_core* TGV_GPU_core* LLT_ROF_GPU_core* CCPiDefines.h +delete PatchSelect_core* Nonlocal_TV_core* shared.h fprintf('%s \n', 'All successfully compiled!'); pathA2 = sprintf(['..' fsep '..' fsep '..' fsep '..' fsep 'demos'], 1i); -cd(pathA2);
\ No newline at end of file +cd(pathA2); diff --git a/src/Matlab/mex_compile/regularisers_CPU/PatchSelect.c b/src/Matlab/mex_compile/regularisers_CPU/PatchSelect.c index d2f6670..1acab29 100644 --- a/src/Matlab/mex_compile/regularisers_CPU/PatchSelect.c +++ b/src/Matlab/mex_compile/regularisers_CPU/PatchSelect.c @@ -52,8 +52,8 @@ void mexFunction( { int number_of_dims, SearchWindow, SimilarWin, NumNeighb; mwSize dimX, dimY, dimZ; - unsigned short *H_i=NULL, *H_j=NULL, *H_k=NULL; - mwSize *dim_array; + const mwSize *dim_array; + unsigned short *H_i=NULL, *H_j=NULL, *H_k=NULL; float *A, *Weights = NULL, h; mwSize dim_array2[3]; /* for 2D data */ mwSize dim_array3[4]; /* for 3D data */ |