diff options
author | Edoardo Pasca <edo.paskino@gmail.com> | 2017-08-04 16:15:17 +0100 |
---|---|---|
committer | Edoardo Pasca <edo.paskino@gmail.com> | 2017-10-11 15:38:09 +0100 |
commit | 36e4c296223f67bb917511089ec59533460f1695 (patch) | |
tree | cfe28281f0a3233b9710ec356b6ebedae32d43df /src/Python | |
parent | 12dbe738d5a2af5573e33a31f1745a50dba165ba (diff) | |
download | regularization-36e4c296223f67bb917511089ec59533460f1695.tar.gz regularization-36e4c296223f67bb917511089ec59533460f1695.tar.bz2 regularization-36e4c296223f67bb917511089ec59533460f1695.tar.xz regularization-36e4c296223f67bb917511089ec59533460f1695.zip |
test facility for regularizers
Diffstat (limited to 'src/Python')
-rw-r--r-- | src/Python/test_regularizers.py | 265 |
1 files changed, 265 insertions, 0 deletions
diff --git a/src/Python/test_regularizers.py b/src/Python/test_regularizers.py new file mode 100644 index 0000000..6abfba4 --- /dev/null +++ b/src/Python/test_regularizers.py @@ -0,0 +1,265 @@ +# -*- coding: utf-8 -*- +""" +Created on Fri Aug 4 11:10:05 2017 + +@author: ofn77899 +""" + +from ccpi.viewer.CILViewer2D import Converter +import vtk + +import regularizers +import matplotlib.pyplot as plt +import numpy as np +import os +from enum import Enum + +class Regularizer(): + '''Class to handle regularizer algorithms to be used during reconstruction + + Currently 5 regularization algorithms are available: + + 1) SplitBregman_TV + 2) FGP_TV + 3) + 4) + 5) + + Usage: + the regularizer can be invoked as object or as static method + Depending on the actual regularizer the input parameter may vary, and + a different default setting is defined. + reg = Regularizer(Regularizer.Algorithm.SplitBregman_TV) + + out = reg(input=u0, regularization_parameter=10., number_of_iterations=30, + tolerance_constant=1e-4, + TV_Penalty=Regularizer.TotalVariationPenalty.l1) + + out2 = Regularizer.SplitBregman_TV(input=u0, regularization_parameter=10., + number_of_iterations=30, tolerance_constant=1e-4, + TV_Penalty=Regularizer.TotalVariationPenalty.l1) + + A number of optional parameters can be passed or skipped + out2 = Regularizer.SplitBregman_TV(input=u0, regularization_parameter=10. ) + + ''' + class Algorithm(Enum): + SplitBregman_TV = regularizers.SplitBregman_TV + FGP_TV = regularizers.FGP_TV + LLT_model = regularizers.LLT_model + # Algorithm + + class TotalVariationPenalty(Enum): + isotropic = 0 + l1 = 1 + # TotalVariationPenalty + + def __init__(self , algorithm): + + self.algorithm = algorithm + self.pars = self.parsForAlgorithm(algorithm) + # __init__ + + def parsForAlgorithm(self, algorithm): + pars = dict() + if algorithm == Regularizer.Algorithm.SplitBregman_TV : + pars['algorithm'] = algorithm + pars['input'] = None + pars['regularization_parameter'] = None + pars['number_of_iterations'] = 35 + pars['tolerance_constant'] = 0.0001 + pars['TV_penalty'] = Regularizer.TotalVariationPenalty.isotropic + elif algorithm == Regularizer.Algorithm.FGP_TV : + pars['algorithm'] = algorithm + pars['input'] = None + pars['regularization_parameter'] = None + pars['number_of_iterations'] = 50 + pars['tolerance_constant'] = 0.001 + pars['TV_penalty'] = Regularizer.TotalVariationPenalty.isotropic + elif algorithm == Regularizer.Algorithm.LLT_model: + pars['algorithm'] = algorithm + pars['input'] = None + pars['regularization_parameter'] = None + pars['time_step'] = None + pars['number_of_iterations'] = None + pars['tolerance_constant'] = None + pars['restrictive_Z_smoothing'] = 0 + + return pars + # parsForAlgorithm + + def __call__(self, input, regularization_parameter, **kwargs): + + if kwargs is not None: + for key, value in kwargs.items(): + #print("{0} = {1}".format(key, value)) + self.pars[key] = value + self.pars['input'] = input + self.pars['regularization_parameter'] = regularization_parameter + #for key, value in self.pars.items(): + # print("{0} = {1}".format(key, value)) + + if self.algorithm == Regularizer.Algorithm.SplitBregman_TV : + return self.algorithm(input, regularization_parameter, + self.pars['number_of_iterations'], + self.pars['tolerance_constant'], + self.pars['TV_penalty'].value ) + elif self.algorithm == Regularizer.Algorithm.FGP_TV : + return self.algorithm(input, regularization_parameter, + self.pars['number_of_iterations'], + self.pars['tolerance_constant'], + self.pars['TV_penalty'].value ) + elif self.algorithm == Regularizer.Algorithm.LLT_model : + #LLT_model(np::ndarray input, double d_lambda, double d_tau, int iter, double d_epsil, int switcher) + # no default + if None in self.pars: + raise Exception("Not all parameters have been provided") + else: + return self.algorithm(input, + regularization_parameter, + self.pars['time_step'] , + self.pars['number_of_iterations'], + self.pars['tolerance_constant'], + self.pars['restrictive_Z_smoothing'] ) + + + # __call__ + + @staticmethod + def SplitBregman_TV(input, regularization_parameter , **kwargs): + reg = Regularizer(Regularizer.Algorithm.SplitBregman_TV) + out = list( reg(input, regularization_parameter, **kwargs) ) + out.append(reg.pars) + return out + + @staticmethod + def FGP_TV(input, regularization_parameter , **kwargs): + reg = Regularizer(Regularizer.Algorithm.FGP_TV) + out = list( reg(input, regularization_parameter, **kwargs) ) + out.append(reg.pars) + return out + + @staticmethod + def LLT_model(input, regularization_parameter , time_step, number_of_iterations, + tolerance_constant, restrictive_Z_smoothing=0): + reg = Regularizer(Regularizer.Algorithm.FGP_TV) + out = list( reg(input, regularization_parameter, time_step=time_step, + number_of_iterations=number_of_iterations, + tolerance_constant=tolerance_constant, + restrictive_Z_smoothing=restrictive_Z_smoothing) ) + out.append(reg.pars) + return out + + +#Example: +# figure; +# Im = double(imread('lena_gray_256.tif'))/255; % loading image +# u0 = Im + .05*randn(size(Im)); u0(u0 < 0) = 0; +# u = SplitBregman_TV(single(u0), 10, 30, 1e-04); + +filename = r"C:\Users\ofn77899\Documents\GitHub\CCPi-FISTA_reconstruction\data\lena_gray_512.tif" +reader = vtk.vtkTIFFReader() +reader.SetFileName(os.path.normpath(filename)) +reader.Update() +#vtk returns 3D images, let's take just the one slice there is as 2D +Im = Converter.vtk2numpy(reader.GetOutput()).T[0]/255 + +#imgplot = plt.imshow(Im) +perc = 0.05 +u0 = Im + (perc* np.random.normal(size=np.shape(Im))) +# map the u0 u0->u0>0 +f = np.frompyfunc(lambda x: 0 if x < 0 else x, 1,1) +u0 = f(u0).astype('float32') + +# plot +fig = plt.figure() +a=fig.add_subplot(2,3,1) +a.set_title('Original') +imgplot = plt.imshow(Im) + +a=fig.add_subplot(2,3,2) +a.set_title('noise') +imgplot = plt.imshow(u0) + + +############################################################################## +# Call regularizer + +####################### SplitBregman_TV ##################################### +# u = SplitBregman_TV(single(u0), 10, 30, 1e-04); + +reg = Regularizer(Regularizer.Algorithm.SplitBregman_TV) + +out = reg(input=u0, regularization_parameter=10., #number_of_iterations=30, + #tolerance_constant=1e-4, + TV_Penalty=Regularizer.TotalVariationPenalty.l1) + +out2 = Regularizer.SplitBregman_TV(input=u0, regularization_parameter=10., number_of_iterations=30, + tolerance_constant=1e-4, + TV_Penalty=Regularizer.TotalVariationPenalty.l1) +out2 = Regularizer.SplitBregman_TV(input=u0, regularization_parameter=10. ) +pars = out2[2] + +a=fig.add_subplot(2,3,3) +a.set_title('SplitBregman_TV') +textstr = 'regularization_parameter=%.2f\niterations=%d\ntolerance=%.2e\npenalty=%s' +textstr = textstr % (pars['regularization_parameter'], + pars['number_of_iterations'], + pars['tolerance_constant'], + pars['TV_penalty'].name) + +# these are matplotlib.patch.Patch properties +props = dict(boxstyle='round', facecolor='wheat', alpha=0.5) +# place a text box in upper left in axes coords +a.text(0.05, 0.95, textstr, transform=a.transAxes, fontsize=14, + verticalalignment='top', bbox=props) +imgplot = plt.imshow(out2[0]) + +###################### FGP_TV ######################################### +# u = FGP_TV(single(u0), 0.05, 100, 1e-04); +out2 = Regularizer.FGP_TV(input=u0, regularization_parameter=0.05, + number_of_iterations=10) +pars = out2[-1] + +a=fig.add_subplot(2,3,4) +a.set_title('FGP_TV') +textstr = 'regularization_parameter=%.2f\niterations=%d\ntolerance=%.2e\npenalty=%s' +textstr = textstr % (pars['regularization_parameter'], + pars['number_of_iterations'], + pars['tolerance_constant'], + pars['TV_penalty'].name) + +# these are matplotlib.patch.Patch properties +props = dict(boxstyle='round', facecolor='wheat', alpha=0.5) +# place a text box in upper left in axes coords +a.text(0.05, 0.95, textstr, transform=a.transAxes, fontsize=14, + verticalalignment='top', bbox=props) +imgplot = plt.imshow(out2[0]) + +###################### LLT_model ######################################### +# * u0 = Im + .03*randn(size(Im)); % adding noise +# [Den] = LLT_model(single(u0), 10, 0.1, 1); +out2 = Regularizer.LLT_model(input=u0, regularization_parameter=10., + time_step=0.1, + tolerance_constant=1e-4, + number_of_iterations=10) +pars = out2[-1] + +a=fig.add_subplot(2,3,5) +a.set_title('LLT_model') +textstr = 'regularization_parameter=%.2f\niterations=%d\ntolerance=%.2e\ntime-step=%f' +textstr = textstr % (pars['regularization_parameter'], + pars['number_of_iterations'], + pars['tolerance_constant'], + pars['time_step'] + ) + +# these are matplotlib.patch.Patch properties +props = dict(boxstyle='round', facecolor='wheat', alpha=0.5) +# place a text box in upper left in axes coords +a.text(0.05, 0.95, textstr, transform=a.transAxes, fontsize=14, + verticalalignment='top', bbox=props) +imgplot = plt.imshow(out2[0]) + + + |