/
configurate_data.py
147 lines (129 loc) · 5.79 KB
/
configurate_data.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
import math
import matplotlib.pyplot as plt
import numpy as np
import os
import tensorflow as tf
from PIL import Image
TRAIN_DATA_SIZE = 120
TEST_DATA_SIZE = 30
IMG_SIZE = 256
OUTPUT_SIZE = 256*256
CATEGORY = 1
FILENAMES = ['../DAGM/Class1_def/',
'../DAGM/Class2_def/',
'../DAGM/Class3_def/',
'../DAGM/Class4_def/',
'../DAGM/Class5_def/',
'../DAGM/Class6_def/']
def savetxt(filename, data):
for i in range(data.shape[0]):
print(filename + ', line ' + str(i))
file = open(filename, 'a')
file.write('{:.9f}'.format(data[i, 0]))
for j in range(1, data.shape[1]):
file.write(',' + '{:.9f}'.format(data[i, j]))
file.write('\n')
file.close()
if __name__ == "__main__":
init = tf.global_variables_initializer()
sess = tf.Session()
with sess.as_default():
if not os.path.exists('./data'):
os.mkdir('./data')
# remove old file
if(os.path.exists('./data/trainImage256.txt')):
os.remove('./data/trainImage256.txt')
if(os.path.exists('./data/trainLABEL' + str(IMG_SIZE) + '.txt')):
os.remove('./data/trainLABEL' + str(IMG_SIZE) + '.txt')
if(os.path.exists('./data/testImage256.txt')):
os.remove('./data/testImage256.txt')
if(os.path.exists('./data/testLABEL' + str(IMG_SIZE) + '.txt')):
os.remove('./data/testLABEL' + str(IMG_SIZE) + '.txt')
cntTrain = 0
cntTest = 0
for n in range(CATEGORY):
for k in range(TRAIN_DATA_SIZE + TEST_DATA_SIZE):
filename = FILENAMES[n] + str(k + 1) + '.png'
print(filename)
imgtf = tf.read_file(filename)
img = tf.image.decode_png(imgtf, channels=1)
resized = tf.image.resize_images(img, [IMG_SIZE, IMG_SIZE], method=tf.image.ResizeMethod.AREA)
array = resized.eval()
if(k < TRAIN_DATA_SIZE):
line = str(cntTrain)
for i in range(IMG_SIZE):
for j in range(IMG_SIZE):
line = line + ',' + str(array[i, j, 0])
line = line + '\n'
file = open('./data/trainImage256.txt', 'a')
file.write(line)
file.close()
cntTrain += 1
else:
line = str(cntTest)
for i in range(IMG_SIZE):
for j in range(IMG_SIZE):
line = line + ',' + str(array[i, j, 0])
line = line + '\n'
file = open('./data/testImage256.txt', 'a')
file.write(line)
file.close()
cntTest += 1
# label #
trnLABEL = []
tstLABEL = []
for n in range(CATEGORY):
x = np.linspace(1.0, 511, IMG_SIZE)
y = np.linspace(1.0, 511, IMG_SIZE)
filename = FILENAMES[n] + 'labels.txt'
label1 = open(filename, 'r')
print('reading ' + filename)
for k in range(TRAIN_DATA_SIZE + TEST_DATA_SIZE):
line = label1.readline()
val = line.split('\t')
num = int(val[0]) - 1
mjr = float(val[1])
mnr = float(val[2])
rot = float(val[3])
cnx = float(val[4])
cny = float(val[5])
# inverse rotate pixels
label = np.zeros([OUTPUT_SIZE*CATEGORY + 1])
label[0] = num # index
for i in range(IMG_SIZE):
for j in range(IMG_SIZE):
dist = math.sqrt((x[i] - cnx)**2 + (y[j] - cny)**2)
xTmp = (x[i] - cnx) * math.cos(-rot) - (y[j] - cny) * math.sin(-rot)
yTmp = (x[i] - cnx) * math.sin(-rot) + (y[j] - cny) * math.cos(-rot)
ang = math.atan2(yTmp, xTmp)
distToEllipse = math.sqrt((mjr * math.cos(ang))**2 + (mnr * math.sin(ang))**2)
if(dist < distToEllipse):
label[(j*IMG_SIZE + i)*CATEGORY + n + 1] = 1.0 # defection
else:
label[(j*IMG_SIZE + i)*CATEGORY + n + 1] = 0.0
# plot test
#if(k == 0):
#plt.figure(figsize=(5, 5))
#z = label[1:OUTPUT_IMG_SIZE*OUTPUT_IMG_SIZE + 1].reshape([OUTPUT_IMG_SIZE, OUTPUT_IMG_SIZE])
#plt.imshow(z)
#plt.show()
if(k < TRAIN_DATA_SIZE):
trnLABEL.append(label)
else:
tstLABEL.append(label)
# normalize
w_array = np.array(trnLABEL)
# for k in range(TRAIN_DATA_SIZE*CATEGORY):
# s = sum(w_array[k, 1:OUTPUT_SIZE*CATEGORY + 1])
# w_array[k, 1:OUTPUT_SIZE*CATEGORY + 1] = w_array[k, 1:OUTPUT_SIZE*CATEGORY + 1]/s
# trnLABEL = w_array.tolist()
w_tst_array = np.array(tstLABEL)
# for k in range(TEST_DATA_SIZE*CATEGORY):
# s = sum(w_tst_array[k, 1:OUTPUT_SIZE*CATEGORY + 1])
# w_tst_array[k, 1:OUTPUT_SIZE*CATEGORY + 1] = w_tst_array[k, 1:OUTPUT_SIZE*CATEGORY + 1]/s
# tstLABEL = w_tst_array.tolist()
#np.savetxt('./data/trainLABEL' + str(IMG_SIZE) + '.txt', trnLABEL, fmt='%.10f', delimiter=',')
#np.savetxt('./data/testLABEL' + str(IMG_SIZE) + '.txt', tstLABEL, fmt='%.10f', delimiter=',')
savetxt('./data/trainLABEL' + str(IMG_SIZE) + '.txt', w_array)
savetxt('./data/testLABEL' + str(IMG_SIZE) + '.txt', w_tst_array)
sess.close()