-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathsrc.py
168 lines (153 loc) · 5.14 KB
/
src.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
from numba import vectorize, guvectorize, float32, int32, jit, cuda
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image
import cv2
import pywt
import math
from scipy import signal
from scipy.stats import entropy
from sklearn.cluster import KMeans
def img_filter(img):
w = 5
wVal = 1/w
conColMatrix = np.zeros(shape=(w,1))
conColMatrix = np.add(conColMatrix, wVal)
conRowMatrix = np.zeros(shape=(1,w))
conRowMatrix = np.add(conRowMatrix, wVal)
a = signal.convolve2d(img, conColMatrix, boundary='symm', mode='same')
return signal.convolve2d(a, conRowMatrix, boundary='symm', mode='same')
@jit(nopython=True)
def T_or_I_of_Neutrsophic(item, minVal, maxVal):
return (item - minVal) / (maxVal - minVal)
@jit(nopython=True)
def EnI_Pixel(item):
return item * math.log2(item)
@jit(nopython=True)
def Alpha_I_Pixel(item,minVal, maxVal):
return (item * minVal) / (maxVal - minVal)
@jit(nopython=True)
def Beta_I_Pixel(item,minVal, maxVal):
return (item * minVal) / (maxVal - minVal)
def Neutrsophic_Image(item):
lmv = img_filter(item)
lmvMin = np.amin(lmv)
lmvMax = np.amax(lmv)
TI_func = np.vectorize(T_or_I_of_Neutrsophic)
T = TI_func(lmv,lmvMin,lmvMax)
absMatrix = np.absolute(np.array(item) - np.array(lmv))
absMatrixMin = np.amin(absMatrix)
absMatrixMax = np.amax(absMatrix)
I = TI_func(absMatrix, absMatrixMin, absMatrixMax)
F = 1 - T
pns = np.zeros((len(T),len(T[0]),3))
pns[:,:,0] = T
pns[:,:,1] = I
pns[:,:,2] = F
return pns
def cal_alpha_beta(img, EnI):
h = len(img)
w = len(img[0])
# EnI_func = np.vectorize(EnI_Pixel)
# EnI = EnI_func(img)
EnMin = 0.0
EnMax = -math.log2(1/h*w)
alphaMin = 0.01
alphaMax = 0.1
alpha = alphaMin + (( alphaMax - alphaMin ) * ( EnI - EnMin ) / ( EnMax - EnMin ))
beta = 1 - alpha
return alpha, beta
def ns(img, enI):
T = img[:,:,0]
I = img[:,:,1]
rowCount = len(T)
colCount = len(T[0])
# T_alpha mean
alphaValT, betaValT = cal_alpha_beta(T, enI)
meanT = img_filter(T)
alphaT = np.ones((rowCount,colCount))
alphaT[I<alphaValT] = T[I<alphaValT]
alphaT[I>=alphaValT] = meanT[I>=alphaValT]
alphaMT = img_filter(alphaT)
# I_alpha mean
meanI = np.absolute(alphaT-alphaMT)
meanImin = np.amin(meanI)
meanImax = np.amax(meanI)
Alpha_I_Pixel_func = np.vectorize(Alpha_I_Pixel)
alphaI = Alpha_I_Pixel_func(meanI,meanImin,meanImax)
EnhT=np.ones((rowCount, colCount))
EnhT[alphaT<=0.5] = 2*(EnhT[alphaT<=0.5] ** 2)
EnhT[alphaT>0.5] = 1-2*(1-EnhT[alphaT>0.5]) ** 2
betaEnhT=np.ones((rowCount, colCount))
betaEnhT[alphaI<betaValT]=alphaT[alphaI<betaValT]
betaEnhT[alphaI>=betaValT]=EnhT[alphaI>=betaValT]
betaT = img_filter(betaEnhT)
betaI = np.absolute(betaEnhT - betaT)
betaImin = np.amin(betaI)
betaImax = np.amax(betaI)
Beta_I_Pixel_func = np.vectorize(Alpha_I_Pixel)
betaEnhI = Beta_I_Pixel_func(betaI,betaImin,betaImax)
value, counts = np.unique(betaEnhI, return_counts=True)
nsEntropy = entropy(counts)
return nsEntropy, alphaValT, betaEnhT, betaEnhI
def get_X_img(img):
enl=0.00001
err=0.0001
while True:
nsEntropy, alpha, betaEnhT, betaEnhI = ns(img, enl)
val = (nsEntropy-enl)/enl
if(val < err):
break
else:
enl = nsEntropy
x = np.zeros((len(img), len(img[0])))
mt = img_filter(betaEnhT)
x[betaEnhI<alpha]=betaEnhT[betaEnhI<alpha]
x[betaEnhI>=alpha]=mt[betaEnhI>=alpha]
return x
img = cv2.imread('C:\\Users\\mfo\\Desktop\\calismalar\\cita_tek.jpg')
luvRawImg = cv2.cvtColor(img, cv2.COLOR_RGB2LUV)
luvImg = np.array(luvRawImg,'float')
grayRawImg = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
grayImg = np.array(grayRawImg,'float')
LL, (LH, HL, HH) = pywt.dwt2(grayImg,'bior2.2')
l,u,v = luvImg[:,:,0], luvImg[:,:,1], luvImg[:,:,2]
# Step 4
LH = cv2.resize(LH,dsize = (len(grayImg[0]), len(grayImg)),interpolation=cv2.INTER_CUBIC)
HL = cv2.resize(HL,dsize = (len(grayImg[0]), len(grayImg)),interpolation=cv2.INTER_CUBIC)
MELH = img_filter(LH)
MEHL = img_filter(HL)
# Step 5
MELH_NS = Neutrsophic_Image(MELH)
MEHL_NS = Neutrsophic_Image(MEHL)
l_NS = Neutrsophic_Image(l)
u_NS = Neutrsophic_Image(u)
v_NS = Neutrsophic_Image(v)
# Step 6,7,8,9
l_NS_X = get_X_img(l_NS)
u_NS_X = get_X_img(u_NS)
v_NS_X = get_X_img(v_NS)
MELH_NS_X = get_X_img(MELH_NS)
MEHL_NS_X = get_X_img(MEHL_NS)
nRows = len(l_NS_X)
nCols = len(l_NS_X[0])
X1 = np.zeros((nRows,nCols,5))
X1[:,:,0] = l_NS_X
X1[:,:,1] = u_NS_X
X1[:,:,2] = v_NS_X
X1[:,:,3] = MELH_NS_X
X1[:,:,4] = MEHL_NS_X
X2 = np.reshape(X1,(nRows*nCols,5))
kmeans = KMeans(n_clusters=4, random_state=0)
clusters = kmeans.fit_predict(X2)
resultImg = clusters.reshape(nRows,nCols)
titles = ['LL', 'LH', 'HL', 'HH', 'LL', 'MELH', 'MEHL', 'HH', 'luvImg', 'resultImg' ]
fig = plt.figure(figsize=(12, 3))
for i, a in enumerate([LL, LH, HL, HH, LL, MELH, MEHL, HH, luvRawImg, resultImg]):
ax = fig.add_subplot(3, 4, i + 1)
ax.imshow(a, interpolation="nearest", cmap=plt.cm.gray)
ax.set_title(titles[i], fontsize=9)
ax.set_xticks([])
ax.set_yticks([])
fig.tight_layout()
plt.show()