-
Notifications
You must be signed in to change notification settings - Fork 0
/
pnp_test.py
277 lines (229 loc) · 12.4 KB
/
pnp_test.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
#!/usr/bin/python
# coding:utf-8
# author: chaidisheng
from __future__ import absolute_import
from __future__ import division
from __future__ import print_function
from utils.utils import batch_PSNR, batch_SSIM
import os
import cv2
import glob
import time
import random
import argparse
import torch
import torch.nn as nn
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import scipy.io as io
from torch.backends import cudnn
from utils.log import init_logging
from Algorithm.ADMM.PnP_ADMM import PnP_ADMM
from Algorithm.ADMM.PnP_DRS import PnP_DRS
from Algorithm.ADMM.PnP_Multi_ADMM import PnP_Multi_ADMM
from Algorithm.ADMM.PnP_Stochastic_ADMM import PnP_Stochastic_ADMM
from Algorithm.ADMM.PnP_Combination_ADMM import PnP_Combination_ADMM
from Algorithm.Half_Quadratic_Splitting.half_quadratic_splitting import PnP_HQS, half_quadratic_splitting_
from Algorithm.Primal_Dual_Splitting.PnP_Pock_Chambolle import PnP_PDS, primal_dual_splitting
from Algorithm.Proximal_Gradient.PnP_FBS import PnP_FBS
from Algorithm.Denoised_Approximate_Message_Passing.PnP_AMP import PnP_AMP
from Algorithm.ADMM.PnP_CEOP import PnP_CEOP
from Algorithm.ADMM.PnP_CE import PnP_CE
from utils.measurement_utils import wgn, SNR
from utils.figure import show_detail, show_double_detail
from utils.measurement_utils import code_diffraction_forward as AA
from utils.measurement_utils import code_diffraction_backward as AAt
from utils.measurement_utils import code_diffraction_backward_complex as Complex_AAt
os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
os.environ["CUDA_VISIBLE_DEVICES"] = "0" # physics specify
def normalize(data):
"""normalize image"""
return data / 255.0
def proj(image, *options):
return np.minimum(np.maximum(image, options[0]), options[1])
def save_mat(tensor):
# save mat file
out = tensor.squeeze().numpy()
image = np.mat(out)
io.savemat('image.mat', {'matrix': image})
def test(args):
# make save direction
print('Start to test on {}'.format(args.test_data))
out_dir = args.save_dir+args.test_data.split('/')[-1]+'/'
if not os.path.exists(out_dir):
os.makedirs(out_dir)
# load data info
print('Loading data info ...\n')
files_source = glob.glob(os.path.join('data', args.test_data, '*.png'))
files_source.sort()
# log info file
logger = init_logging(out_dir)
logger.info("Algorithm parameters are {}".format(args))
# process data
name, PSNR, SSIM, TIME = [], [], [], []
averaged_psnr, averaged_ssim, averaged_time = (0, 0, 0)
for files in files_source:
Img = cv2.imread(files)
Img = normalize(np.float32(Img[:, :, 0]))
original_image = Img
Img = np.expand_dims(Img, 0)
Img = np.expand_dims(Img, 1)
ISource = torch.Tensor(Img).to(args.device)
# initial array
# x_init = 0.5 * torch.ones_like(ISource)
x_init = torch.rand_like(ISource)
# forward operator or backward operator
A = lambda x: AA(x, sampling_rate=args.sampling_rate, seed=args.random_seed)
At = lambda x: AAt(x, x_init.shape, sampling_rate=args.sampling_rate)
Complex_At = lambda x: Complex_AAt(x, x_init.shape, sampling_rate=args.sampling_rate)
# save images file name
filename = files.split('/')[-1].split('.')[0] # get the name of image file
print("\n%s image of phase retrieval is begin!" % filename)
hist_logdir = out_dir+filename+'_hist'
name.append(filename)
# set noise model
pre_measure = torch.sqrt(torch.sum(A(ISource[:]) ** 2, dim=3, keepdim=True))
gaussian_noise, sigma_w = wgn(ISource, args.snr, seed=args.random_seed)
measure = pre_measure+gaussian_noise.view(pre_measure.shape) # 5.65
print('SNR is: {0:.2f}'.format(SNR(pre_measure, measure).item()))
# DnCnn: blind: 0.01=25.5, non-blind: 0.004=16
# options = dict(lambd=0.01, max_iters=50, beta=0.5, verbose=True)
# start_time = time.time()
# out_image = PnP_ADMM(measure[:], A, At, 'DnCNN', x_init, **options)
# stop_time = time.time()
# options = dict(lambd=[0.001, 0.005], NUM_CHANNELS=2, max_iters=50, verbose=True)
# out_image = PnP_Multi_ADMM(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(lambd=[0.01, 0.01], max_iters=50, beta=0.5, verbose=True)
# out_image = PnP_Stochastic_ADMM(measure[:], A, At, ['DnCNN', 'BM3D'], x_init, **options)
# options = dict(lambd=[0.001, 0.005], max_iters=50, verbose=True)
# out_image = PnP_Combination_ADMM(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(lambd=[0.003, 0.01], beta=0.5, numbers=3, max_iters=50, verbose=True)
# out_image = PnP_CEOP(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(lambd=0.01, max_iters=50, verbose=True)
# out_image = PnP_DRS(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(lambd=0.01, rho_scale=1.0, max_iters=1, max_inner_iters=70, verbose=True)
# out_image = PnP_HQS(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(lambd=0.01, max_iters=50, verbose=True)
# out_image = PnP_PDS(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(max_iters=30, tol=1e-7, sigma_w=sigma_w, sigma_hat=[10./255, 20./255], record_objective=False,
# adaptive=False, accelerate=False, backtrack=False, restart=False)
# out_image = PnP_FBS(measure[:], A, At, 'DnCNN', x_init, **options)
# options = dict(max_iters=50, eps=1e1, blind=False, denoiser='DnCNN')
# out_image = PnP_AMP(measure[:], A, Complex_At, x_init, **options)
# Algorithm parameters
# args.sigma[0] if sigma_w < 0.10 else args.sigma[1]
# options = dict(lambd=args.sigma, max_iters=args.max_iters,
# beta=args.beta, verbose=args.verbose, logdir=hist_logdir, ori_img=ISource,)
# options = dict(sigma_hat=args.sigma, max_iters=args.max_iters, tol=1e-7, record_objective=False,
# adaptive=False, accelerate=False, backtrack=False, restart=False, logdir=hist_logdir,
# verbose=args.verbose, ori_img=ISource,)
options = dict(lambd=torch.tensor(args.sigma), max_iters=args.max_iters, ori_img=ISource,
noise_level=sigma_w, beta=args.beta, verbose=args.verbose,)
PnP_Method = dict(PnP_ADMM=PnP_ADMM, PnP_FBS=PnP_FBS, PnP_CE=PnP_CE)
try:
start_time = time.time()
out_image = PnP_Method[args.PnP_Method](measure[:], A, At, args.noise_type,
args.denoiser, x_init, **options)
stop_time = time.time()
except KeyError:
print("PnP Method Fault!")
test_time = stop_time-start_time
TIME.append(test_time)
averaged_time += test_time
psnr, ssim = batch_PSNR(out_image, ISource, 1.), batch_SSIM(out_image, ISource, 1.)
PSNR.append(psnr)
SSIM.append(ssim)
averaged_psnr += psnr
averaged_ssim += ssim
print("%s PSNR: %.2fdb, SSIM: %.2f, Consuming Time: % .2fs. " % (files, psnr, ssim, test_time))
# figure = plt.figure("restoration image")
plt.figure("restoration image")
plt.subplot(1, 2, 1)
plt.imshow(torch.squeeze(ISource).cpu(), cmap='gray')
plt.title("inf db")
plt.subplot(1, 2, 2)
plt.imshow(torch.squeeze(out_image).cpu(), cmap='gray')
plt.title(str(format(psnr, '.2f'))+'db')
plt.show()
# plt.figure("concatenate image")
# plt.imshow(np.concatenate((torch.squeeze(ISource), torch.squeeze(out_image)), axis=1), cmap='gray')
# plt.title("Original-Restoration")
# plt.show()
org_dir = out_dir+str(files.split('/')[2].split('.')[0])+'.png'
logdir = out_dir+str(files.split('/')[2].split('.')[0])+'-'+str(format(psnr, '.2f'))+'db'+'.png'
logdir_double = out_dir+str(files.split('/')[2].split('.')[0])+'_'+str(format(psnr, '.2f'))+'db'+'.png'
list_ = [np.expand_dims(original_image, 2),
np.expand_dims(proj(torch.squeeze(out_image).cpu().numpy(), 0., 1.), 2)]
width, length = *[[40, 100] if ISource.size()[3] == 256 else [180, 150]], \
*[[100, 160] if ISource.size()[3] == 256 else [240, 250]]
xy = [160, 40] if ISource.size()[3] == 256 else [250, 180]
yx = [120, 136] if ISource.size()[3] == 256 else [200, 392]
show_detail(np.expand_dims(original_image, 2), org_dir, width, length, 2, 1, 0, xy, yx)
show_double_detail(logdir_double, width, length, 2, 1, 0, xy, yx, *list_)
show_detail(np.expand_dims(proj(out_image.squeeze().cpu().numpy(), 0., 1.), 2),
logdir, width, length, 2, 1, 0, xy, yx)
name.append("Average")
averaged_psnr /= len(files_source)
PSNR.append(averaged_psnr)
averaged_ssim /= len(files_source)
SSIM.append(averaged_ssim)
averaged_time /= len(files_source)
TIME.append(averaged_time)
print("\nOn the test data Average PSNR: %.2fdb, Average SSIM: %.2f, Average Consuming Time: %.2fs. " %
(averaged_psnr, averaged_ssim, averaged_time))
pd.DataFrame({'name': np.array(name), 'psnr': np.round(np.array(PSNR), 2), 'ssim': np.round(np.array(SSIM), 2),
'test_time': np.round(np.array(TIME), 2)}).to_csv(out_dir+'/metrics.csv', index=True)
if __name__ == "__main__":
# python3 train_full_realsn.py --sigma=15./255 or python3 train_full_realsn.py --sigma 15./255
parser = argparse.ArgumentParser(description="PnP_Test")
parser.add_argument("--PnP_Method", type=str, default='PnP_CE', help='PnP_ADMM or PnP_FBS or PnP_CE')
parser.add_argument("--save_dir", type=str, default="./result/PnP_CE_blind_25/", help='result of log files')
parser.add_argument("--test_data", type=str, default='TestImages', help='test on Set12 or TestImages')
parser.add_argument("--sigma", nargs='+', type=float, default=[15./255, 25./255],
help='noise level of used denoiser')
parser.add_argument("--denoiser", type=str, default="DnCNN", help='method of used denoiser')
parser.add_argument("--noise_type", type=bool, default=True, help='noise type of used denoiser')
# seed = random.randint(1, 1000)
parser.add_argument("--random_seed", type=int, default=0, help='random seed of measurements')
parser.add_argument("--sampling_rate", type=int, default=1, help='number of measurements')
parser.add_argument("--snr", type=float, default=25, help='signal-to-noise ratio')
parser.add_argument("--max_iters", type=int, default=60, help='max iterations')
parser.add_argument("--beta", type=float, default=0.5, help='averaged operator parameter')
parser.add_argument("--verbose", type=bool, default=True, help='log info running')
parser.add_argument("--gpu", type=int, default=0, help='GPU index')
opt = parser.parse_args()
# show software version
print("PyTorch version : ", torch.__version__) # PyTorch version
print("CUDA version : ", torch.version.cuda) # Corresponding CUDA version
print("cuDNN version : ", torch.backends.cudnn.version()) # Corresponding cuDNN version
if torch.cuda.is_available():
print("Cuda current device : ", torch.cuda.current_device()) # Corresponding CUDA serial number
print("GPU type : ", torch.cuda.get_device_name(0)) # GPU type
# specify gpu number
if opt.gpu is None:
raise ValueError("please provide gpu id")
opt.device = torch.device("cuda:%d" % opt.gpu if torch.cuda.is_available() else "cpu") # logic specify
if torch.cuda.is_available():
torch.backends.cudnn.enabled = False # default: False
torch.backends.cudnn.benchmark = False # default: False
torch.backends.cudnn.deterministic = True # default: True
print('CUDA available: {}'.format(torch.cuda.is_available()))
# clear cache of gpu
torch.cuda.empty_cache()
# specify random seed
random.seed(0)
np.random.seed(0)
torch.manual_seed(0)
torch.cuda.manual_seed(0)
torch.cuda.manual_seed_all(0)
if torch.cuda.is_available():
print("Returns the current %d random seed of the current GPU." % torch.cuda.initial_seed())
# print parameters
print("\n### running PnP model ###")
print("> Parameters:")
for p, v in zip(opt.__dict__.keys(), opt.__dict__.values()):
print('\t{}: {}'.format(p, v))
print('\n')
# test model
test(opt)