-
Notifications
You must be signed in to change notification settings - Fork 34
/
utils.py
235 lines (184 loc) · 7.4 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
234
235
# -*- coding: utf-8 -*-
"""
License: MIT
@author: gaj
E-mail: anjing_guo@hnu.edu.cn
"""
import cv2
import numpy as np
from scipy import ndimage
from scipy import signal
import scipy.misc as misc
def upsample_bilinear(image, ratio):
h,w,c = image.shape
re_image = cv2.resize(image, (w*ratio, h*ratio), interpolation=cv2.INTER_LINEAR)
return re_image
def upsample_bicubic(image, ratio):
h,w,c = image.shape
re_image = cv2.resize(image, (w*ratio, h*ratio), interpolation=cv2.INTER_CUBIC)
return re_image
def upsample_interp23(image, ratio):
image = np.transpose(image, (2, 0, 1))
b,r,c = image.shape
CDF23 = 2*np.array([0.5, 0.305334091185, 0, -0.072698593239, 0, 0.021809577942, 0, -0.005192756653, 0, 0.000807762146, 0, -0.000060081482])
d = CDF23[::-1]
CDF23 = np.insert(CDF23, 0, d[:-1])
BaseCoeff = CDF23
first = 1
for z in range(1,np.int(np.log2(ratio))+1):
I1LRU = np.zeros((b, 2**z*r, 2**z*c))
if first:
I1LRU[:, 1:I1LRU.shape[1]:2, 1:I1LRU.shape[2]:2]=image
first = 0
else:
I1LRU[:,0:I1LRU.shape[1]:2,0:I1LRU.shape[2]:2]=image
for ii in range(0,b):
t = I1LRU[ii,:,:]
for j in range(0,t.shape[0]):
t[j,:]=ndimage.correlate(t[j,:],BaseCoeff,mode='wrap')
for k in range(0,t.shape[1]):
t[:,k]=ndimage.correlate(t[:,k],BaseCoeff,mode='wrap')
I1LRU[ii,:,:]=t
image = I1LRU
re_image=np.transpose(I1LRU, (1, 2, 0))
return re_image
def upsample_mat_interp23(image, ratio=4):
'''2 pixel shift compare with original matlab version'''
shift=2
h,w,c = image.shape
basecoeff = np.array([[-4.63495665e-03, -3.63442646e-03, 3.84904063e-18,
5.76678319e-03, 1.08358664e-02, 1.01980790e-02,
-9.31747402e-18, -1.75033181e-02, -3.17660068e-02,
-2.84531643e-02, 1.85181518e-17, 4.42450253e-02,
7.71733386e-02, 6.70554910e-02, -2.85299239e-17,
-1.01548683e-01, -1.78708388e-01, -1.60004642e-01,
3.61741232e-17, 2.87940558e-01, 6.25431459e-01,
8.97067600e-01, 1.00107877e+00, 8.97067600e-01,
6.25431459e-01, 2.87940558e-01, 3.61741232e-17,
-1.60004642e-01, -1.78708388e-01, -1.01548683e-01,
-2.85299239e-17, 6.70554910e-02, 7.71733386e-02,
4.42450253e-02, 1.85181518e-17, -2.84531643e-02,
-3.17660068e-02, -1.75033181e-02, -9.31747402e-18,
1.01980790e-02, 1.08358664e-02, 5.76678319e-03,
3.84904063e-18, -3.63442646e-03, -4.63495665e-03]])
coeff = np.dot(basecoeff.T, basecoeff)
I1LRU = np.zeros((ratio*h, ratio*w, c))
I1LRU[shift::ratio, shift::ratio, :]=image
for i in range(c):
temp = I1LRU[:, :, i]
temp = ndimage.convolve(temp, coeff, mode='wrap')
I1LRU[:, :, i]=temp
return I1LRU
def gaussian2d (N, std):
t=np.arange(-(N-1)/2,(N+2)/2)
t1,t2=np.meshgrid(t,t)
std=np.double(std)
w = np.exp(-0.5*(t1/std)**2)*np.exp(-0.5*(t2/std)**2)
return w
def kaiser2d (N, beta):
t=np.arange(-(N-1)/2,(N+2)/2)/np.double(N-1)
t1,t2=np.meshgrid(t,t)
t12=np.sqrt(t1*t1+t2*t2)
w1=np.kaiser(N,beta)
w=np.interp(t12,t,w1)
w[t12>t[-1]]=0
w[t12<t[0]]=0
return w
def fir_filter_wind(Hd,w):
"""
compute fir filter with window method
Hd: desired freqeuncy response (2D)
w: window (2D)
"""
hd=np.rot90(np.fft.fftshift(np.rot90(Hd,2)),2)
h=np.fft.fftshift(np.fft.ifft2(hd))
h=np.rot90(h,2)
h=h*w
h=h/np.sum(h)
return h
def downgrade_images(I_MS, I_PAN, ratio, sensor=None):
"""
downgrade MS and PAN by a ratio factor with given sensor's gains
"""
I_MS=np.double(I_MS)
I_PAN=np.double(I_PAN)
I_MS = np.transpose(I_MS, (2, 0, 1))
I_PAN = np.squeeze(I_PAN)
ratio=np.double(ratio)
flag_PAN_MTF=0
if sensor=='QB':
flag_resize_new = 2
GNyq = np.asarray([0.34, 0.32, 0.30, 0.22],dtype='float32') # Band Order: B,G,R,NIR
GNyqPan = 0.15
elif sensor=='IKONOS':
flag_resize_new = 2 #MTF usage
GNyq = np.asarray([0.26,0.28,0.29,0.28],dtype='float32') # Band Order: B,G,R,NIR
GNyqPan = 0.17;
elif sensor=='GeoEye1':
flag_resize_new = 2 # MTF usage
GNyq = np.asarray([0.23,0.23,0.23,0.23],dtype='float32') # Band Order: B,G,R,NIR
GNyqPan = 0.16
elif sensor=='WV2':
flag_resize_new = 2 # MTF usage
GNyq = [0.35,0.35,0.35,0.35,0.35,0.35,0.35,0.27]
GNyqPan = 0.11
elif sensor=='WV3':
flag_resize_new = 2 #MTF usage
GNyq = 0.29 * np.ones(8)
GNyqPan = 0.15
else:
'''the default way'''
flag_resize_new = 1
'''the default downgrading method is gaussian'''
if flag_resize_new == 1:
# I_MS_LP = np.zeros((I_MS.shape[0],int(np.round(I_MS.shape[1]/ratio)+ratio),int(np.round(I_MS.shape[2]/ratio)+ratio)))
#
# for idim in range(I_MS.shape[0]):
# imslp_pad=np.pad(I_MS[idim,:,:],int(2*ratio),'symmetric')
# I_MS_LP[idim,:,:]=misc.imresize(imslp_pad,1/ratio,'bicubic',mode='F')
#
# I_MS_LR = I_MS_LP[:,2:-2,2:-2]
#
# I_PAN_pad=np.pad(I_PAN,int(2*ratio),'symmetric')
# I_PAN_LR=misc.imresize(I_PAN_pad,1/ratio,'bicubic',mode='F')
# I_PAN_LR=I_PAN_LR[2:-2,2:-2]
sig = (1/(2*(2.772587)/ratio**2))**0.5
kernel = np.multiply(cv2.getGaussianKernel(9, sig), cv2.getGaussianKernel(9,sig).T)
t=[]
for i in range(I_MS.shape[0]):
temp = signal.convolve2d(I_MS[i, :, :], kernel, mode='same', boundary = 'wrap')
temp = temp[0::int(ratio), 0::int(ratio)]
temp = np.expand_dims(temp, 0)
t.append(temp)
I_MS_LR = np.concatenate(t, axis=0)
I_PAN_LR = signal.convolve2d(I_PAN, kernel, mode='same', boundary = 'wrap')
I_PAN_LR = I_PAN_LR[0::int(ratio), 0::int(ratio)]
elif flag_resize_new==2:
N=41
I_MS_LP=np.zeros(I_MS.shape)
fcut=1/ratio
for j in range(I_MS.shape[0]):
#fir filter with window method
alpha = np.sqrt(((N-1)*(fcut/2))**2/(-2*np.log(GNyq[j])))
H=gaussian2d(N,alpha)
Hd=H/np.max(H)
w=kaiser2d(N,0.5)
h=fir_filter_wind(Hd,w)
I_MS_LP[j,:,:] = ndimage.filters.correlate(I_MS[j,:,:],np.real(h),mode='nearest')
if flag_PAN_MTF==1:
#fir filter with window method
alpha = np.sqrt(((N-1)*(fcut/2))**2/(-2*np.log(GNyqPan)))
H=gaussian2d(N,alpha)
Hd=H/np.max(H)
h=fir_filter_wind(Hd,w)
I_PAN = ndimage.filters.correlate(I_PAN,np.real(h),mode='nearest')
I_PAN_LR=I_PAN[int(ratio/2):-1:int(ratio),int(ratio/2):-1:int(ratio)]
else:
#bicubic resize
I_PAN_pad=np.pad(I_PAN,int(2*ratio),'symmetric')
I_PAN_LR=misc.imresize(I_PAN_pad,1/ratio,'bicubic',mode='F')
I_PAN_LR=I_PAN_LR[2:-2,2:-2]
I_MS_LR=I_MS_LP[:,int(ratio/2):-1:int(ratio),int(ratio/2):-1:int(ratio)]
I_MS_LR = np.transpose(I_MS_LR, (1, 2, 0))
I_PAN_LR = np.expand_dims(I_PAN_LR, -1)
return I_MS_LR,I_PAN_LR