Merge pull request #52 from vais-ral/NonlDiffusion

Nonlinear diffusion module

7 files changed, 427 insertions, 21 deletions

diff --git a/Wrappers/Python/ccpi/filters/regularisers.py b/Wrappers/Python/ccpi/filters/regularisers.py
index e6814e8..eec8c4d 100644
--- a/Wrappers/Python/ccpi/filters/regularisers.py
+++ b/Wrappers/Python/ccpi/filters/regularisers.py
@@ -2,8 +2,8 @@
 script which assigns a proper device core function based on a flag ('cpu' or 'gpu')
 """
 
-from ccpi.filters.cpu_regularisers_cython import TV_ROF_CPU, TV_FGP_CPU, TV_SB_CPU, dTV_FGP_CPU, TNV_CPU 
-from ccpi.filters.gpu_regularisers import TV_ROF_GPU, TV_FGP_GPU, TV_SB_GPU, dTV_FGP_GPU
+from ccpi.filters.cpu_regularisers_cython import TV_ROF_CPU, TV_FGP_CPU, TV_SB_CPU, dTV_FGP_CPU, TNV_CPU, NDF_CPU
+from ccpi.filters.gpu_regularisers import TV_ROF_GPU, TV_FGP_GPU, TV_SB_GPU, dTV_FGP_GPU, NDF_GPU
 
 def ROF_TV(inputData, regularisation_parameter, iterations,
                      time_marching_parameter,device='cpu'):
@@ -91,3 +91,22 @@ def TNV(inputData, regularisation_parameter, iterations, tolerance_param):
                      regularisation_parameter,
                      iterations, 
                      tolerance_param)
+def NDF(inputData, regularisation_parameter, edge_parameter, iterations,
+                     time_marching_parameter, penalty_type, device='cpu'):
+    if device == 'cpu':
+        return NDF_CPU(inputData,
+                     regularisation_parameter,
+                     edge_parameter,
+                     iterations, 
+                     time_marching_parameter,
+                     penalty_type)
+    elif device == 'gpu':
+        return NDF_GPU(inputData,
+                     regularisation_parameter,
+                     edge_parameter,
+                     iterations, 
+                     time_marching_parameter,
+                     penalty_type)
+    else:
+        raise ValueError('Unknown device {0}. Expecting gpu or cpu'\
+                         .format(device))
diff --git a/Wrappers/Python/demos/demo_cpu_regularisers.py b/Wrappers/Python/demos/demo_cpu_regularisers.py
index 7443b83..3567f91 100644
--- a/Wrappers/Python/demos/demo_cpu_regularisers.py
+++ b/Wrappers/Python/demos/demo_cpu_regularisers.py
@@ -12,7 +12,7 @@ import matplotlib.pyplot as plt
 import numpy as np
 import os
 import timeit
-from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV, TNV
+from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV, TNV, NDF
 from qualitymetrics import rmse
 ###############################################################################
 def printParametersToString(pars):
@@ -190,11 +190,58 @@ plt.title('{}'.format('CPU results'))
 
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("_____________FGP-dTV (2D)__________________")
+print ("________________NDF (2D)___________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
 fig = plt.figure(4)
+plt.suptitle('Performance of NDF regulariser using the CPU')
+a=fig.add_subplot(1,2,1)
+a.set_title('Noisy Image')
+imgplot = plt.imshow(u0,cmap="gray")
+
+# set parameters
+pars = {'algorithm' : NDF, \
+        'input' : u0,\
+        'regularisation_parameter':0.06, \
+        'edge_parameter':0.04,\
+        'number_of_iterations' :1000 ,\
+        'time_marching_parameter':0.025,\
+        'penalty_type':1
+        }
+        
+print ("#############NDF CPU################")
+start_time = timeit.default_timer()
+ndf_cpu = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'],'cpu')  
+             
+rms = rmse(Im, ndf_cpu)
+pars['rmse'] = rms
+
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,2,2)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_cpu, cmap="gray")
+plt.title('{}'.format('CPU results'))
+
+
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+print ("_____________FGP-dTV (2D)__________________")
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+
+## plot 
+fig = plt.figure(5)
 plt.suptitle('Performance of FGP-dTV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -247,7 +294,7 @@ print ("__________Total nuclear Variation__________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(5)
+fig = plt.figure(6)
 plt.suptitle('Performance of TNV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -321,7 +368,7 @@ print ("_______________ROF-TV (3D)_________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(6)
+fig = plt.figure(7)
 plt.suptitle('Performance of ROF-TV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy 15th slice of a volume')
@@ -361,7 +408,7 @@ print ("_______________FGP-TV (3D)__________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(7)
+fig = plt.figure(8)
 plt.suptitle('Performance of FGP-TV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -410,7 +457,7 @@ print ("_______________SB-TV (3D)_________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(8)
+fig = plt.figure(9)
 plt.suptitle('Performance of SB-TV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -451,13 +498,58 @@ a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
 imgplot = plt.imshow(sb_cpu3D[10,:,:], cmap="gray")
 plt.title('{}'.format('Recovered volume on the CPU using SB-TV'))
 
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+print ("________________NDF (3D)___________________")
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+
+## plot 
+fig = plt.figure(10)
+plt.suptitle('Performance of NDF regulariser using the CPU')
+a=fig.add_subplot(1,2,1)
+a.set_title('Noisy volume')
+imgplot = plt.imshow(noisyVol[10,:,:],cmap="gray")
+
+# set parameters
+pars = {'algorithm' : NDF, \
+        'input' : noisyVol,\
+        'regularisation_parameter':0.06, \
+        'edge_parameter':0.04,\
+        'number_of_iterations' :1000 ,\
+        'time_marching_parameter':0.025,\
+        'penalty_type':  1
+        }
+        
+print ("#############NDF CPU################")
+start_time = timeit.default_timer()
+ndf_cpu3D = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'])  
+             
+rms = rmse(idealVol, ndf_cpu3D)
+pars['rmse'] = rms
+
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,2,2)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_cpu3D[10,:,:], cmap="gray")
+plt.title('{}'.format('Recovered volume on the CPU using NDF iterations'))
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 print ("_______________FGP-dTV (3D)__________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(9)
+fig = plt.figure(11)
 plt.suptitle('Performance of FGP-dTV regulariser using the CPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
diff --git a/Wrappers/Python/demos/demo_cpu_vs_gpu_regularisers.py b/Wrappers/Python/demos/demo_cpu_vs_gpu_regularisers.py
index d8e2da7..05db23e 100644
--- a/Wrappers/Python/demos/demo_cpu_vs_gpu_regularisers.py
+++ b/Wrappers/Python/demos/demo_cpu_vs_gpu_regularisers.py
@@ -12,7 +12,7 @@ import matplotlib.pyplot as plt
 import numpy as np
 import os
 import timeit
-from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV
+from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV, NDF
 from qualitymetrics import rmse
 ###############################################################################
 def printParametersToString(pars):
@@ -306,11 +306,98 @@ else:
 
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("____________FGP-dTV bench___________________")
+print ("_______________NDF bench___________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
 fig = plt.figure(4)
+plt.suptitle('Comparison of NDF regulariser using CPU and GPU implementations')
+a=fig.add_subplot(1,4,1)
+a.set_title('Noisy Image')
+imgplot = plt.imshow(u0,cmap="gray")
+
+# set parameters
+pars = {'algorithm' : NDF, \
+        'input' : u0,\
+        'regularisation_parameter':0.06, \
+        'edge_parameter':0.04,\
+        'number_of_iterations' :1000 ,\
+        'time_marching_parameter':0.025,\
+        'penalty_type':  1
+        }
+        
+print ("#############NDF CPU####################")
+start_time = timeit.default_timer()
+ndf_cpu = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'],'cpu')
+             
+rms = rmse(Im, ndf_cpu)
+pars['rmse'] = rms
+
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,4,2)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_cpu, cmap="gray")
+plt.title('{}'.format('CPU results'))
+
+
+print ("##############NDF GPU##################")
+start_time = timeit.default_timer()
+ndf_gpu = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'],'gpu')
+             
+rms = rmse(Im, ndf_gpu)
+pars['rmse'] = rms
+pars['algorithm'] = NDF
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,4,3)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_gpu, cmap="gray")
+plt.title('{}'.format('GPU results'))
+
+print ("--------Compare the results--------")
+tolerance = 1e-05
+diff_im = np.zeros(np.shape(rof_cpu))
+diff_im = abs(ndf_cpu - ndf_gpu)
+diff_im[diff_im > tolerance] = 1
+a=fig.add_subplot(1,4,4)
+imgplot = plt.imshow(diff_im, vmin=0, vmax=1, cmap="gray")
+plt.title('{}'.format('Pixels larger threshold difference'))
+if (diff_im.sum() > 1):
+    print ("Arrays do not match!")
+else:
+    print ("Arrays match")
+
+
+
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+print ("____________FGP-dTV bench___________________")
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+
+## plot 
+fig = plt.figure(5)
 plt.suptitle('Comparison of FGP-dTV regulariser using CPU and GPU implementations')
 a=fig.add_subplot(1,4,1)
 a.set_title('Noisy Image')
diff --git a/Wrappers/Python/demos/demo_gpu_regularisers.py b/Wrappers/Python/demos/demo_gpu_regularisers.py
index 25d8d85..b873700 100644
--- a/Wrappers/Python/demos/demo_gpu_regularisers.py
+++ b/Wrappers/Python/demos/demo_gpu_regularisers.py
@@ -12,7 +12,7 @@ import matplotlib.pyplot as plt
 import numpy as np
 import os
 import timeit
-from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV
+from ccpi.filters.regularisers import ROF_TV, FGP_TV, SB_TV, FGP_dTV, NDF
 from qualitymetrics import rmse
 ###############################################################################
 def printParametersToString(pars):
@@ -50,7 +50,7 @@ u0 = u0.astype('float32')
 u_ref = u_ref.astype('float32')
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("____________ROF-TV bench___________________")
+print ("____________ROF-TV regulariser_____________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
@@ -92,7 +92,7 @@ plt.title('{}'.format('GPU results'))
 
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("____________FGP-TV bench___________________")
+print ("____________FGP-TV regulariser_____________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
@@ -141,7 +141,7 @@ plt.title('{}'.format('GPU results'))
 
 
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("____________SB-TV bench___________________")
+print ("____________SB-TV regulariser______________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
@@ -186,12 +186,60 @@ a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
 imgplot = plt.imshow(sb_gpu, cmap="gray")
 plt.title('{}'.format('GPU results'))
 
+
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
-print ("____________FGP-dTV bench___________________")
+print ("_______________NDF regulariser_____________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
 fig = plt.figure(4)
+plt.suptitle('Performance of the NDF regulariser using the GPU')
+a=fig.add_subplot(1,2,1)
+a.set_title('Noisy Image')
+imgplot = plt.imshow(u0,cmap="gray")
+
+# set parameters
+pars = {'algorithm' : NDF, \
+        'input' : u0,\
+        'regularisation_parameter':0.06, \
+        'edge_parameter':0.04,\
+        'number_of_iterations' :1000 ,\
+        'time_marching_parameter':0.025,\
+        'penalty_type':  1
+        }
+
+print ("##############NDF GPU##################")
+start_time = timeit.default_timer()
+ndf_gpu = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'],'gpu')  
+             
+rms = rmse(Im, ndf_gpu)
+pars['rmse'] = rms
+pars['algorithm'] = NDF
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,2,2)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_gpu, cmap="gray")
+plt.title('{}'.format('GPU results'))
+
+
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+print ("____________FGP-dTV bench___________________")
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+
+## plot 
+fig = plt.figure(5)
 plt.suptitle('Performance of the FGP-dTV regulariser using the GPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -266,7 +314,7 @@ print ("_______________ROF-TV (3D)_________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(5)
+fig = plt.figure(6)
 plt.suptitle('Performance of ROF-TV regulariser using the GPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy 15th slice of a volume')
@@ -306,7 +354,7 @@ print ("_______________FGP-TV (3D)__________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(6)
+fig = plt.figure(7)
 plt.suptitle('Performance of FGP-TV regulariser using the GPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -354,7 +402,7 @@ print ("_______________SB-TV (3D)__________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(7)
+fig = plt.figure(8)
 plt.suptitle('Performance of SB-TV regulariser using the GPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
@@ -395,12 +443,60 @@ a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
 imgplot = plt.imshow(sb_gpu3D[10,:,:], cmap="gray")
 plt.title('{}'.format('Recovered volume on the GPU using SB-TV'))
 
+
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+print ("_______________NDF-TV (3D)_________________")
+print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
+
+## plot 
+fig = plt.figure(9)
+plt.suptitle('Performance of NDF regulariser using the GPU')
+a=fig.add_subplot(1,2,1)
+a.set_title('Noisy Image')
+imgplot = plt.imshow(noisyVol[10,:,:],cmap="gray")
+
+# set parameters
+pars = {'algorithm' : NDF, \
+        'input' : noisyVol,\
+        'regularisation_parameter':0.06, \
+        'edge_parameter':0.04,\
+        'number_of_iterations' :1000 ,\
+        'time_marching_parameter':0.025,\
+        'penalty_type':  1
+        }
+
+print ("#############NDF GPU####################")
+start_time = timeit.default_timer()
+ndf_gpu3D = NDF(pars['input'], 
+              pars['regularisation_parameter'],
+              pars['edge_parameter'], 
+              pars['number_of_iterations'],
+              pars['time_marching_parameter'], 
+              pars['penalty_type'],'gpu')
+
+rms = rmse(idealVol, ndf_gpu3D)
+pars['rmse'] = rms
+
+txtstr = printParametersToString(pars)
+txtstr += "%s = %.3fs" % ('elapsed time',timeit.default_timer() - start_time)
+print (txtstr)
+a=fig.add_subplot(1,2,2)
+
+# these are matplotlib.patch.Patch properties
+props = dict(boxstyle='round', facecolor='wheat', alpha=0.75)
+# place a text box in upper left in axes coords
+a.text(0.15, 0.25, txtstr, transform=a.transAxes, fontsize=14,
+         verticalalignment='top', bbox=props)
+imgplot = plt.imshow(ndf_gpu3D[10,:,:], cmap="gray")
+plt.title('{}'.format('Recovered volume on the GPU using NDF'))
+
+
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 print ("_______________FGP-dTV (3D)________________")
 print ("%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%")
 
 ## plot 
-fig = plt.figure(8)
+fig = plt.figure(10)
 plt.suptitle('Performance of FGP-dTV regulariser using the GPU')
 a=fig.add_subplot(1,2,1)
 a.set_title('Noisy Image')
diff --git a/Wrappers/Python/setup-regularisers.py.in b/Wrappers/Python/setup-regularisers.py.in
index 0681cc4..b900efe 100644
--- a/Wrappers/Python/setup-regularisers.py.in
+++ b/Wrappers/Python/setup-regularisers.py.in
@@ -37,6 +37,7 @@ extra_include_dirs += [os.path.join(".." , ".." , "Core"),
                        os.path.join(".." , ".." , "Core",  "regularisers_GPU" , "TV_FGP" ) , 
                        os.path.join(".." , ".." , "Core",  "regularisers_GPU" , "TV_ROF" ) , 
                        os.path.join(".." , ".." , "Core",  "regularisers_GPU" , "TV_SB" ) ,
+                       os.path.join(".." , ".." , "Core",  "regularisers_GPU" , "NDF" ) ,
                        os.path.join(".." , ".." , "Core",  "regularisers_GPU" , "dTV_FGP" ) , 
 						   "."]
 
diff --git a/Wrappers/Python/src/cpu_regularisers.pyx b/Wrappers/Python/src/cpu_regularisers.pyx
index abbf3b0..7ed8fa1 100644
--- a/Wrappers/Python/src/cpu_regularisers.pyx
+++ b/Wrappers/Python/src/cpu_regularisers.pyx
@@ -21,10 +21,10 @@ cimport numpy as np
 cdef extern float TV_ROF_CPU_main(float *Input, float *Output, float lambdaPar, int iterationsNumb, float tau, int dimX, int dimY, int dimZ);
 cdef extern float TV_FGP_CPU_main(float *Input, float *Output, float lambdaPar, int iterationsNumb, float epsil, int methodTV, int nonneg, int printM, int dimX, int dimY, int dimZ);
 cdef extern float SB_TV_CPU_main(float *Input, float *Output, float lambdaPar, int iterationsNumb, float epsil, int methodTV, int printM, int dimX, int dimY, int dimZ);
+cdef extern float Diffusion_CPU_main(float *Input, float *Output, float lambdaPar, float sigmaPar, int iterationsNumb, float tau, int penaltytype, int dimX, int dimY, int dimZ);
 cdef extern float TNV_CPU_main(float *Input, float *u, float lambdaPar, int maxIter, float tol, int dimX, int dimY, int dimZ);
 cdef extern float dTV_FGP_CPU_main(float *Input, float *InputRef, float *Output, float lambdaPar, int iterationsNumb, float epsil, float eta, int methodTV, int nonneg, int printM, int dimX, int dimY, int dimZ);
 
-
 #****************************************************************#
 #********************** Total-variation ROF *********************#
 #****************************************************************#
@@ -275,3 +275,47 @@ def TNV_3D(np.ndarray[np.float32_t, ndim=3, mode="c"] inputData,
     # Run TNV iterations for 3D (X,Y,Channels) data 
     TNV_CPU_main(&inputData[0,0,0], &outputData[0,0,0], regularisation_parameter, iterationsNumb, tolerance_param, dims[2], dims[1], dims[0])
     return outputData
+#****************************************************************#
+#***************Nonlinear (Isotropic) Diffusion******************#
+#****************************************************************#
+def NDF_CPU(inputData, regularisation_parameter, edge_parameter, iterationsNumb,time_marching_parameter, penalty_type):
+    if inputData.ndim == 2:
+        return NDF_2D(inputData, regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type)
+    elif inputData.ndim == 3:
+        return NDF_3D(inputData, regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type)
+
+def NDF_2D(np.ndarray[np.float32_t, ndim=2, mode="c"] inputData, 
+                     float regularisation_parameter,
+                     float edge_parameter,
+                     int iterationsNumb,                     
+                     float time_marching_parameter,
+                     int penalty_type):
+    cdef long dims[2]
+    dims[0] = inputData.shape[0]
+    dims[1] = inputData.shape[1]
+    
+    cdef np.ndarray[np.float32_t, ndim=2, mode="c"] outputData = \
+            np.zeros([dims[0],dims[1]], dtype='float32')   
+    
+    # Run Nonlinear Diffusion iterations for 2D data 
+    Diffusion_CPU_main(&inputData[0,0], &outputData[0,0], regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type, dims[0], dims[1], 1)    
+    return outputData
+            
+def NDF_3D(np.ndarray[np.float32_t, ndim=3, mode="c"] inputData, 
+                     float regularisation_parameter,
+                     float edge_parameter,
+                     int iterationsNumb,                     
+                     float time_marching_parameter,
+                     int penalty_type):
+    cdef long dims[3]
+    dims[0] = inputData.shape[0]
+    dims[1] = inputData.shape[1]
+    dims[2] = inputData.shape[2]
+    
+    cdef np.ndarray[np.float32_t, ndim=3, mode="c"] outputData = \
+            np.zeros([dims[0],dims[1],dims[2]], dtype='float32')
+    
+    # Run Nonlinear Diffusion iterations for  3D data 
+    Diffusion_CPU_main(&inputData[0,0,0], &outputData[0,0,0], regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type, dims[2], dims[1], dims[0])
+
+    return outputData
diff --git a/Wrappers/Python/src/gpu_regularisers.pyx b/Wrappers/Python/src/gpu_regularisers.pyx
index 36eec95..b0775054 100644
--- a/Wrappers/Python/src/gpu_regularisers.pyx
+++ b/Wrappers/Python/src/gpu_regularisers.pyx
@@ -21,6 +21,7 @@ cimport numpy as np
 cdef extern void TV_ROF_GPU_main(float* Input, float* Output, float lambdaPar, int iter, float tau, int N, int M, int Z);
 cdef extern void TV_FGP_GPU_main(float *Input, float *Output, float lambdaPar, int iter, float epsil, int methodTV, int nonneg, int printM, int N, int M, int Z);
 cdef extern void TV_SB_GPU_main(float *Input, float *Output, float lambdaPar, int iter, float epsil, int methodTV, int printM, int N, int M, int Z);
+cdef extern void NonlDiff_GPU_main(float *Input, float *Output, float lambdaPar, float sigmaPar, int iterationsNumb, float tau, int penaltytype, int N, int M, int Z);
 cdef extern void dTV_FGP_GPU_main(float *Input, float *InputRef, float *Output, float lambdaPar, int iterationsNumb, float epsil, float eta, int methodTV, int nonneg, int printM, int N, int M, int Z);
 
 # Total-variation Rudin-Osher-Fatemi (ROF)
@@ -114,6 +115,27 @@ def dTV_FGP_GPU(inputData,
                      methodTV,
                      nonneg,
                      printM)
+# Nonlocal Isotropic Diffusion (NDF)
+def NDF_GPU(inputData,
+                     regularisation_parameter,
+                     edge_parameter,
+                     iterations, 
+                     time_marching_parameter,
+                     penalty_type):
+    if inputData.ndim == 2:
+        return NDF_GPU_2D(inputData,
+                     regularisation_parameter,
+                     edge_parameter,
+                     iterations, 
+                     time_marching_parameter,
+                     penalty_type)
+    elif inputData.ndim == 3:
+        return NDF_GPU_3D(inputData,
+                     regularisation_parameter,
+                     edge_parameter,
+                     iterations, 
+                     time_marching_parameter,
+                     penalty_type)                     
 #****************************************************************#
 #********************** Total-variation ROF *********************#
 #****************************************************************#
@@ -336,3 +358,48 @@ def FGPdTV3D(np.ndarray[np.float32_t, ndim=3, mode="c"] inputData,
                        printM,
                        dims[2], dims[1], dims[0]);
     return outputData 
+
+#****************************************************************#
+#***************Nonlinear (Isotropic) Diffusion******************#
+#****************************************************************#
+def NDF_GPU_2D(np.ndarray[np.float32_t, ndim=2, mode="c"] inputData, 
+                     float regularisation_parameter,
+                     float edge_parameter,
+                     int iterationsNumb,                     
+                     float time_marching_parameter,
+                     int penalty_type):
+    cdef long dims[2]
+    dims[0] = inputData.shape[0]
+    dims[1] = inputData.shape[1]
+    
+    cdef np.ndarray[np.float32_t, ndim=2, mode="c"] outputData = \
+            np.zeros([dims[0],dims[1]], dtype='float32')
+    
+    #rangecheck = penalty_type < 1 and penalty_type > 3
+    #if not rangecheck:
+#        raise ValueError('Choose penalty type as 1 for Huber, 2 - Perona-Malik, 3 - Tukey Biweight')
+    
+    # Run Nonlinear Diffusion iterations for 2D data 
+    # Running CUDA code here  
+    NonlDiff_GPU_main(&inputData[0,0], &outputData[0,0], regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type, dims[0], dims[1], 1)    
+    return outputData
+            
+def NDF_GPU_3D(np.ndarray[np.float32_t, ndim=3, mode="c"] inputData, 
+                     float regularisation_parameter,
+                     float edge_parameter,
+                     int iterationsNumb,                     
+                     float time_marching_parameter,
+                     int penalty_type):
+    cdef long dims[3]
+    dims[0] = inputData.shape[0]
+    dims[1] = inputData.shape[1]
+    dims[2] = inputData.shape[2]
+    
+    cdef np.ndarray[np.float32_t, ndim=3, mode="c"] outputData = \
+            np.zeros([dims[0],dims[1],dims[2]], dtype='float32')    
+       
+    # Run Nonlinear Diffusion iterations for  3D data 
+    # Running CUDA code here  
+    NonlDiff_GPU_main(&inputData[0,0,0], &outputData[0,0,0], regularisation_parameter, edge_parameter, iterationsNumb, time_marching_parameter, penalty_type, dims[2], dims[1], dims[0])
+
+    return outputData