-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutils.py
233 lines (182 loc) · 6.89 KB
/
utils.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
import tensorflow as tf
import numpy as np
import random
import math
import h5py
import cv2
import os
def time_calculate(sec):
s = sec % 60
m = sec // 60
h = m // 60
m = m % 60
return h, m, s
def str2int_list(str):
num_list = str.split(',')
for idx, elem in enumerate(num_list):
num_list[idx] = int(elem)
return num_list
def scale2str(scale):
scale_str = ''
for elem in scale:
scale_str = scale_str + str(elem) + 'x'
return scale_str
def show_all_variables():
model_vars = tf.trainable_variables()
tf.contrib.slim.model_analyzer.analyze_vars(model_vars, print_info=True)
''' For Load Datasets '''
def load_291(scale):
label = []
data = []
for elem in scale:
filename = '291_' + str(elem) + 'x.h5'
file_dir = os.path.join(os.getcwd(), 'data', 'Train', filename)
with h5py.File(file_dir, 'r') as f:
label.append(list(f['label']))
data.append(list(f['data']))
data = np.concatenate([d for d in data], axis=0)
label = np.concatenate([l for l in label], axis=0)
label, data = _shuffle((label, data))
return label, data
def load_set5(scale=2, color_space='y'):
GT_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Set5', 'gt')
ILR_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Set5', 'bicubic_' + str(scale) + 'x')
label = []
data = []
for img in os.listdir(GT_DIR):
IMG_PATH = os.path.join(GT_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
label.append(read_img[:,:,2:3])
else:
label.append(read_img)
for img in os.listdir(ILR_DIR):
IMG_PATH = os.path.join(ILR_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
data.append(read_img[:,:,2:3])
else:
data.append(read_img)
return _normalize(label), _normalize(data)
def load_set14(scale=2, color_space='y'):
GT_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Set14', 'gt')
ILR_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Set14', 'bicubic_' + str(scale) + 'x')
label = []
data = []
for img in os.listdir(GT_DIR):
IMG_PATH = os.path.join(GT_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
label.append(read_img[:,:,2:3])
else:
label.append(read_img)
for img in os.listdir(ILR_DIR):
IMG_PATH = os.path.join(ILR_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
data.append(read_img[:,:,2:3])
else:
data.append(read_img)
return _normalize(label), _normalize(data)
def load_b100(scale=2, color_space='y'):
GT_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'B100', 'gt')
ILR_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'B100', 'bicubic_' + str(scale) + 'x')
label = []
data = []
for img in os.listdir(GT_DIR):
IMG_PATH = os.path.join(GT_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
label.append(read_img[:,:,2:3])
else:
label.append(read_img)
for img in os.listdir(ILR_DIR):
IMG_PATH = os.path.join(ILR_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
data.append(read_img[:,:,2:3])
else:
data.append(read_img)
return _normalize(label), _normalize(data)
def load_urban100(scale=2, color_space='y'):
GT_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Urban100', 'gt')
ILR_DIR = os.path.join(os.getcwd(), 'data', 'Test', 'Urban100', 'bicubic_' + str(scale) + 'x')
label = []
data = []
for img in os.listdir(GT_DIR):
IMG_PATH = os.path.join(GT_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
label.append(read_img[:,:,2:3])
else:
label.append(read_img)
for img in os.listdir(ILR_DIR):
IMG_PATH = os.path.join(ILR_DIR, img)
read_img = cv2.imread(IMG_PATH, cv2.IMREAD_COLOR)
if color_space == 'y' or color_space == 'Y':
data.append(read_img[:,:,2:3])
else:
data.append(read_img)
return _normalize(label), _normalize(data)
''' For Dataset Pre-processing'''
def create_sub_patches(x, patch_size=41, stride=41):
(HR, LR) = x
HR_patches = []
LR_patches = []
for idx in range(len(HR)):
HR_img = HR[idx]
LR_img = LR[idx]
row = HR_img.shape[0]
col = HR_img.shape[1]
row_cnt = row//patch_size
col_cnt = col//patch_size
for i in range(row_cnt):
for j in range(col_cnt):
HR_img_crop = HR_img[i*patch_size:(i+1)*patch_size, j*patch_size:(j+1)*patch_size]
LR_img_crop = LR_img[i*patch_size:(i+1)*patch_size, j*patch_size:(j+1)*patch_size]
HR_patches.append(HR_img_crop)
LR_patches.append(LR_img_crop)
if col - col_cnt*patch_size > col/3:
HR_img_crop = HR_img[i*patch_size:(i+1)*patch_size, -patch_size:]
LR_img_crop = LR_img[i*patch_size:(i+1)*patch_size, -patch_size:]
HR_patches.append(HR_img_crop)
LR_patches.append(LR_img_crop)
if row - row_cnt * patch_size > patch_size/3:
for j in range(col_cnt):
HR_img_crop = HR_img[-patch_size:, j*patch_size:(j+1)*patch_size]
LR_img_crop = LR_img[-patch_size:, j*patch_size:(j+1)*patch_size]
HR_patches.append(HR_img_crop)
LR_patches.append(LR_img_crop)
if col - col_cnt*patch_size > patch_size/3:
HR_img_crop = HR_img[-patch_size:, -patch_size:]
LR_img_crop = LR_img[-patch_size:, -patch_size:]
HR_patches.append(HR_img_crop)
LR_patches.append(LR_img_crop)
return HR_patches, LR_patches
def _shuffle(x):
(HR, LR) = x
seed = 777
np.random.seed(seed)
np.random.shuffle(HR)
np.random.seed(seed)
np.random.shuffle(LR)
return HR, LR
def _normalize(x):
return np.array(x) / 255.
def denormalize(x):
x *= 255.
x = np.clip(x, 0, 255)
x = x.astype('uint8')
return np.array(x)
''' Invert Color Channel '''
def bgr2ycrcb(img):
cvt_img = cv2.cvtColor(x[idx], cv2.COLOR_BGR2YCrCb)
return cvt_img
def ycrcb2bgr(img):
cvt_img = cv2.cvtColor(img, cv2.COLOR_YCrCb2BGR)
return cvt_img
if __name__ == '__main__' :
label, data = load_291([2])
print(np.array(label).shape)
print(np.array(data).shape)
print(scale2str([2,3,4]))