-
Notifications
You must be signed in to change notification settings - Fork 0
/
Noise_and_filter.py
167 lines (154 loc) · 5.06 KB
/
Noise_and_filter.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
import cv2
import random
import itertools
import copy
import numpy
def show_img(name,img):
cv2.namedWindow(name)#cv2.WINDOW_NORMAL可以手动改变大小
cv2.imshow(name,img)
cv2.waitKey(0)
cv2.destroyAllWindows()
class Noise(object):#噪声
def read(self,path):#读入要处理的图片
self.original_img=cv2.imread(path)
self.img=cv2.cvtColor(self.original_img,cv2.COLOR_BGR2GRAY)
show_img('The original image',self.img)
def salt_and_pepper_noise(self,SNR):#增加椒盐噪声
im=copy.copy(self.img)
s=im.shape
ns=int((1-SNR)*s[0]*s[1])
random_list=list(itertools.product(range(s[0]),range(s[1])))
random_coordinate=random.sample(random_list,ns)#生成随机的不重复的二维坐标
for i in random_coordinate:
r=random.randint(0,1)
if r==0:
im[i[0]][i[1]]=0
else:
im[i[0]][i[1]]=255
cv2.imwrite('salt_and_pepper_noise.jpg',im)
show_img('Slat and perpper noise',im)
return im
def gaussian_noise(self,Xmean,sigma):#增加高斯噪声
im=copy.copy(self.img)
s=im.shape
for i in range(s[0]):
for j in range(s[1]):
r=int(random.gauss(Xmean,sigma))
im[i][j]=im[i][j]+r
if im[i][j]>255:
im[i][j]=255
elif im[i][j]<0:
im[i][j]=0
cv2.imwrite('Gaussian noise.jpg',im)
show_img('Gaussian noise',im)
def impluse_noise(self,SNR,type):#可选择脉冲噪声类型 type={'white','black'}
im=copy.copy(self.img)
s=im.shape
ns=int((1-SNR)*s[0]*s[1])
random_list=list(itertools.product(range(s[0]),range(s[1])))
random_coordinate=random.sample(random_list,ns)#生成随机的不重复的二维坐标
if type=='white':
for i in random_coordinate:
im[i[0]][i[1]]=255
cv2.imwrite('Impluse noise(White).jpg',im)
show_img('Impluse noise(White)',im)
else:
for i in random_coordinate:
im[i[0]][i[1]]=0
cv2.imwrite('Impluse noise(Black).jpg',im)
show_img('Impluse noise(Black)',im)
class Filter(object):#滤波
def __init__(self,img):
s=img.shape
print(s)
#if s[2] == 3:
#self.bgr_img=img
#else:
self.grey_img=img
def median_filter(self,Kernel):#中值滤波
im=cv2.medianBlur(self.grey_img,Kernel)
show_img('Median filter',im)
cv2.imwrite('Median filter.jpg',im)
return(im)
def mean_filter(self,m,n):#均值滤波
im=cv2.blur(self.grey_img,(m,n))
show_img('Mean filter',im)
cv2.imwrite('Mean filter.jpg',im)
return(im)
def grey_histogram_equalization(self):#灰色图直方图均衡化
gray_statistics={}#像素的灰度值统计
normalized_statistis={}#每一个像素值得频率计算
translate_table={}#像素转换表
new_img=copy.copy(img)
for i in range(256):
gray_statistics[i]=0
normalized_statistis[i]=0
s=self.grey_img.shape
for i in range(s[0]):
for j in range(s[1]):
gray_statistics[self.grey_img[i][j]]+=1
total=s[0]*s[1]
for i in gray_statistics:
normalized_statistis[i]=gray_statistics[i]/total
for i in range(256):
translate_table[i]=round(255*self.sum(normalized_statistis,i))
for i in range(s[0]):
for j in range(s[1]):
new_img[i][j]=translate_table[self.grey_img[i][j]]
show_img('Histogram Equalization',new_img)
cv2.imwrite('Histogram Equalization.jpg',im)
def bgr_histogram_equalization(self):#因为考虑到bgr三个维度，所以比灰度要再增加两个维度，其他步骤相同
bgr_statistics=[{},{},{}]
normalized_statistics=[{},{},{}]
translate_table=[{},{},{}]
new_img=copy.copy(self.bgr_img)
for m in bgr_statistics:
for i in range(256):
m[i]=0
print('The initial bgr_statistics has done!')
for n in normalized_statistics:
for i in range(256):
n[i]=0
print('The initial normalized_statistics has done!')
s=self.bgr_img.shape
total=s[0]*s[1]
for i in range(s[0]):
for j in range(s[1]):
for index in range(3):
bgr_statistics[index][self.bgr_img[i][j][index]]+=1
print('(%d,%d) bgr_statistics has done!' % (i,j))
print('The bgr_statistics has done!!')
for index in range(3):
for i in range(256):
normalized_statistics[index][i]=bgr_statistics[index][i]/total
print('normalized_statistics[%d] has done!' % index)
for index in range(3):
for i in range(256):
translate_table[index][i]=round(255*self.sum(normalized_statistics[index],i))
print('translate_table[%d] has done!' % index)
for i in range(s[0]):
for j in range(s[1]):
for index in range(3):
new_img[i][j][index]=translate_table[index][self.bgr_img[i][j][index]]
print(new_img[i][j])
show_img('Bgr Histogram Equalization',new_img)
cv2.imwrite('Bgr Histogram Equalization.jpg',new_img)
def sum(self,dict,num):#求和函数
total=0
for i in range(num):
total+=dict[i]
return total
'''
im=Noise()
im.read('00367.png')
img=im.gaussian_noise(0,45)
img=im.impluse_noise(0.8,'white')
img=im.impluse_noise(0.8,'black')
img=im.salt_and_pepper_noise(0.8)
img=cv2.imread('landscape.jpg')
f=Filter(img)
f.median_filter(5)
f.mean_filter(3,5)
f.grey_histogram_equalization()
f.bgr_histogram_equalization()
'''