-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathgetColorExact.py
130 lines (112 loc) · 3.08 KB
/
getColorExact.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
import numpy as np
from scipy.sparse import coo_matrix
from scipy.sparse.linalg import spsolve
import math
from colorSpace import ntsc2rgb
import pdb
import pickle
def getColorExact(colorIm,ntscIm):
n = np.size(ntscIm,0)
m = np.size(ntscIm,1)
imgSize = n*m
nI = np.zeros(ntscIm.shape)
nI[:,:,0] = ntscIm[:,:,0]
indsM = np.arange(0,imgSize).reshape((n,m),order='F')
print "indsM=\n"
print indsM
test = np.ravel(colorIm,order='F')
test = test.nonzero()
lblInds = np.array(test)
lblInds = lblInds.transpose()
print "lblInds.shape=\n"
print lblInds.shape
wd = 1;
len = 0;
consts_len = 0;
col_inds = np.zeros((imgSize*(2*wd+1)**2,1),float)
row_inds = np.zeros((imgSize*(2*wd+1)**2,1),float)
vals = np.zeros((imgSize*(2*wd+1)**2,1),float)
gvals = np.zeros((1,(2*wd+1)**2))
colorIm = colorIm.astype(int)
#pdb.set_trace()
for j in range(m):
for i in range(n):
if (~colorIm[i,j]):
tlen = 0;
ii = max(0,i-wd)
while (ii<=min(i+wd,n-1)):
jj = max(0,j-wd)
while (jj<=min(j+wd,m-1)):
if (ii!=i) or (jj!=j):
row_inds[len] = consts_len
col_inds[len] = indsM[ii,jj]
gvals[0,tlen] = ntscIm[ii,jj,0]
len=len+1
tlen=tlen+1
jj = jj + 1
ii = ii + 1
t_val = ntscIm[i,j,0]
gvals[0,tlen] = t_val
c_var = np.mean((gvals[0,0:tlen+1] - np.mean(gvals[0,0:tlen+1]))**2)
csig = c_var*0.6
mgv = np.min((gvals[0,0:tlen]-t_val)**2)
if (csig<(-mgv/math.log(0.01))):
csig = -mgv/math.log(0.01)
if (csig<0.000002):
csig = 0.000002
gvals[0,0:tlen] = np.exp(-(((gvals[0,0:tlen]-t_val)**2)/csig))
gvals[0,0:tlen] = gvals[0,0:tlen]/np.sum(gvals[0,0:tlen])
vals[(len-1)-(tlen-1):len,0] = -(gvals[0,0:tlen].T)
row_inds[len] = consts_len
col_inds[len] = indsM[i,j]
vals[len] = 1
consts_len = consts_len + 1
vals = vals[0:len]
col_inds = col_inds[0:len]
row_inds = row_inds[0:len]
#print "vals=\n" + vals
#print "col_inds=\n" + col_inds
#print "row_inds=\n" + row_inds
vals = vals.transpose()
vals.shape = (np.size(vals,1),)
col_inds = col_inds.transpose()
col_inds.shape = (np.size(col_inds,1),)
row_inds = row_inds.transpose()
row_inds.shape = (np.size(row_inds,1),)
print "row_inds.max="
print row_inds.max()
print "consts_len="
print consts_len
A = coo_matrix((vals,(row_inds,col_inds)),shape=(consts_len,imgSize)).tocsc()
#A = np.squeeze(np.asarray(A))
#A.shape = (consts_len,imgSize)
#print type(A)
#print A
#print A.shape
b = np.zeros((consts_len,1))
#print "A=\n" + A
#print "b=\n" + b
with open('imgcolor.pkl','wb') as f:
pickle.dump(vals,f)
pickle.dump(row_inds,f)
pickle.dump(col_inds,f)
pickle.dump(consts_len,f)
pickle.dump(imgSize,f)
pickle.dump(b,f)
t = 1
while (t<=2):
curIm = ntscIm[:,:,t]
for i in range(np.size(lblInds,0)):
b = np.ravel(b,order='F')
curIm = np.ravel(curIm,order='F')
b[lblInds[i]]=curIm[lblInds[i]]
b = b.reshape((consts_len,1),order='F')
new_vals = spsolve(A,b)
print "Pass"
print t
print "Compeleted"
nI[:,:,t] = new_vals.reshape((n,m),order='F')
t = t+1
snI = nI
#nI = ntsc2rgb(nI)
return snI