-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathmain.py
167 lines (131 loc) · 5.67 KB
/
main.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
import rawpy
import cv2
import argparse
import numpy as np
from scipy import signal
def adjust_gamma(image, gamma):
invGamma = 1.0 / gamma
table = np.array([((i / 255.0) ** invGamma) * 255
for i in np.arange(0, 256)]).astype("uint8")
return cv2.LUT(image, table)
def getRaw(path):
raw = rawpy.imread(path)
return raw.raw_image_visible
def colorFilterArray (data):
shape = data.shape
img = np.zeros([(shape[0]), (shape[1]), 3], dtype=np.double)
for i in range(0, shape[0], 2):
for j in range(0, shape[1], 2):
img [i+1, j+1, 0] = np.double( data[i+1, j+1]) #B
img [i+1, j, 1] = np.double(data[i+1, j]) #G
img [i, j+1, 1] = np.double(data[i, j+1]) #G
img [i, j, 2] = np.double(data[i, j]) #R /65535.0
img = img*(1/65535.0)
return img
def demosaicing(img, typeDemosaic):
return bilinearDemosaicingOpt(img)
def bilinearDemosaicingOpt(img):
maskG = np.array([[0, 0.25, 0], [0.25, 0, 0.25], [0, 0.25, 0]])
maskRB = np.array([[0.25, 0.5, 0.25], [0.5, 0, 0.5], [0.25, 0.5, 0.25]])
b = cv2.filter2D(img[:,:,0], -1, maskRB)
g = cv2.filter2D(img[:,:,1], -1,maskG)
r = cv2.filter2D(img[:,:,2], -1, maskRB)
shape = img.shape
#Recovery original values
for i in range(0, shape[0]-2, 2):
for j in range(0, shape[1]-2, 2):
b [i+1, j+1] = img[i+1, j+1, 0] #B
g [i+1, j] = img[i+1, j, 1] #G
g [i, j+1] = img[i, j+1, 1] #G
r [i, j] = img[i, j, 2] #R /65535.0
return cv2.merge((b,g,r))
def bilinearDemosaicing(img):
imgFinal = img.copy()
shape = img.shape
for i in range(1, shape[0]-2):
for j in range(1, shape[1]-2):
if (img[i, j, 1] == 0):
imgFinal[i, j, 1] = (img[i, j+1, 1] + img[i, j-1, 1] + img[i+1, j, 1] + img[i-1, j, 1])*0.25
if (img[i, j, 2] == 0):
if (img[i, j+1, 2] != 0 or img[i, j-1, 2] != 0):
imgFinal[i, j, 2] = (img[i, j+1, 2] + img[i, j-1, 2])*0.5
elif(img[i+1, j, 2] != 0 or img[i-1, j, 2] != 0):
imgFinal[i, j, 2] = (img[i+1, j, 2] + img[i-1, j, 2])*0.5
else:
imgFinal[i, j, 2] = (img[i-1, j+1, 2] + img[i+1, j-1, 2] + img[i-1, j-1, 2] + img[i+1, j+1, 2])*0.25
if (img[i, j, 0] == 0):
if (img[i, j-1, 0] != 0 or img[i, j+1, 0] != 0):
imgFinal[i, j, 0] = (img[i, j+1, 0] + img[i, j-1, 0])*0.5
elif(img[i+1, j, 0] != 0 or img[i-1, j, 2] != 0):
imgFinal[i, j, 0] = (img[i+1, j, 0] + img[i-1, j, 0])*0.5
else:
imgFinal[i, j, 0] = (img[i-1, j+1, 0] + img[i+1, j-1, 0] + img[i-1, j-1, 0] + img[i+1, j+1, 0])*0.25
return imgFinal[2:shape[0]-2, 2:shape[1]-2, :]
def balance_channel(channel, cutoff):
low = np.percentile(channel, cutoff)
high = np.percentile(channel, 100 - cutoff)
new_channel = np.uint8(np.clip((channel-low)*255.0/(high - low), 0, 255))
return new_channel
def automaticWhiteBalance(img, cutoff):
b = balance_channel(img[:,:,0], cutoff)
g = balance_channel(img[:,:,1], cutoff)
r = balance_channel(img[:,:,2], cutoff)
return cv2.merge((b,g,r))
def manualWhiteBalance (img):
posX = 2307
posY = 2289
valR = img[posY, posX, 2]
valG = img[posY, posX, 1]
valB = img[posY, posX, 0]
newChannelB = np.uint8(np.clip((img[:,:,0]*255.0/valB), 0, 255))
newChannelG = np.uint8(np.clip((img[:,:,1]*255.0/valG), 0, 255))
newChannelR = np.uint8(np.clip((img[:,:,2]*255.0/valR), 0, 255))
return cv2.merge((newChannelB, newChannelG, newChannelR))
def whitePatch(img):
maxR = np.max(img[:,:,2])
maxB = np.max(img[:,:,0])
maxG = np.max(img[:,:,1])
while(maxG < 0.45):
img[:,:,1] = img[:,:,1] * 2
maxG = np.max(img[:,:,1])
alpha = maxG/maxR
beta = maxG/maxB
newChannelB = np.uint8(np.clip((img[:,:,0]*255.0*beta), 0, 255))
newChannelG = np.uint8(np.clip((img[:,:,1]*255.0), 0, 255))
newChannelR = np.uint8(np.clip((img[:,:,2]*255.0*alpha), 0, 255))
return cv2.merge((newChannelB, newChannelG, newChannelR))
def grayWorld(img):
maxG = np.max(img[:,:,1])
while(maxG < 0.45):
img[:,:,1] = img[:,:,1] * 2
maxG = np.max(img[:,:,1])
avgR = np.mean(img[:,:,2])
avgB = np.mean(img[:,:,0])
avgG = np.mean(img[:,:,1])
alpha = avgG/avgR
beta = avgG/avgB
newChannelB = np.uint8(np.clip((img[:,:,0]*255.0*beta), 0, 255))
newChannelG = np.uint8(np.clip((img[:,:,1]*255.0), 0, 255))
newChannelR = np.uint8(np.clip((img[:,:,2]*255.0*alpha), 0, 255))
return cv2.merge((newChannelB, newChannelG, newChannelR))
def whiteBalance(img, typeWhiteBalance):
if (typeWhiteBalance == "manual"):
return manualWhiteBalance(img)
elif (typeWhiteBalance == "whitePatch"):
return whitePatch(img)
elif (typeWhiteBalance == "grayWorld"):
return grayWorld(img)
else:
return automaticWhiteBalance(img, 1.0)
parser = argparse.ArgumentParser()
parser.add_argument("-path", type=str)
parser.add_argument("-gamma", default=2.2, type=float)
parser.add_argument("-whiteBalance", default="automatic", type=str)
parser.add_argument("-demosaic", default="bilinear", type=str)
args = parser.parse_args()
bayerData = getRaw(args.path)
img = colorFilterArray(bayerData)
imgFinal = demosaicing(img, args.demosaic)
imgFinalWhiteBalance = whiteBalance(imgFinal, args.whiteBalance) #(balance_white(imgFinal, 2))
cv2.imwrite("foo.png", imgFinalWhiteBalance)
cv2.imwrite("foo2.png", adjust_gamma(imgFinalWhiteBalance, args.gamma))