-
Notifications
You must be signed in to change notification settings - Fork 0
/
image_process.py
171 lines (137 loc) · 5.6 KB
/
image_process.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
169
170
171
import cv2
import numpy as np
from matplotlib import pyplot as plt
from scipy import fftpack
import os
import re
default_hist_param = 30
default_dft_param = 4
default_dct_param = 4
default_grad_param = 5
default_scale_param = 35
histogram = 'get_histogram'
dft = 'get_dft'
dct = 'get_dct'
gradient = 'get_gradient'
scale = 'get_scale'
class Image():
def __init__(self, image_path, test=False):
self.image_path = image_path
self._label_regex = re.compile(r'(?<=s)\d+')
#self._compiled_label_number = int(re.findall(self._label_regex, self.image_path)[0])
self.label_number = int(re.findall(self._label_regex, self.image_path)[0])
'''
if test:
self.is_tested = True
else:
self.is_tested = False
'''
number_regex = re.compile(r'\d+(?=\.pgm)')
self.order_number = re.findall(number_regex, image_path)[0]
self.matrix = cv2.imread(image_path, 0)
'''
def __setattr__(self, arg, new_value):
if arg == 'is_tested' and new_value == False:
if self.label_number is None or self.label_number == 0:
self.label_number = self._compiled_label_number
self.__dict__['is_tested'] = False
elif arg == 'is_tested' and new_value == True:
if self.label_number is not None or self.label_number == 0:
self.label_number = None
self.__dict__['is_tested'] = True
else:
self.__dict__[arg] = new_value
'''
def get_histogram(self, cols_number=default_hist_param):
hist, _ = np.histogram(self.matrix, bins=np.linspace(0, 255, cols_number))
return hist
def get_dft(self, mat_size=default_dft_param):
f = np.fft.fft2(self.matrix)
f = f[0:mat_size, 0:mat_size]
f = cv2.normalize(np.abs(f), None, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX)
return f
def get_dct(self, mat_size=default_dct_param):
c = fftpack.dct(self.matrix, axis=1)
c = fftpack.dct(c, axis=0)
c = c[0:mat_size, 0:mat_size]
c = cv2.normalize(c, None, alpha=0, beta=255, norm_type=cv2.NORM_MINMAX)
return c
def get_gradient(self, window_height=default_grad_param):
rows_number = self.matrix.shape[0]
i = 0
features_number = rows_number - 2*window_height
result = np.zeros(features_number)
while i < features_number:
upper_window = self.matrix[i:i+window_height, :]
lower_window = self.matrix[i+window_height:i+2*window_height]
result[i] = cv2.norm(upper_window, lower_window)
i += 1
return result
def get_scale(self, scale=default_scale_param):
h, w = self.matrix.shape
new_size = (int(w * (scale / 100)), int(h * (scale / 100)))
scaled_matrix = cv2.resize(self.matrix, new_size, interpolation=cv2.INTER_CUBIC)
return scaled_matrix
def show_features(self, cols_number=default_hist_param, dft_mat_size=default_dft_param,
dct_mat_size=default_dct_param, window_height=default_grad_param,
scale=default_scale_param):
fig, [[ax1, ax2, ax3], [ax4, ax5, _]] = plt.subplots(2, 3, figsize=(15, 15))
#print(ax)
hist = self.get_histogram(cols_number)
ax1.bar(range(len(hist)), hist)
ax1.set_xlabel('columns')
ax1.set_ylabel('frequency')
ticks = len(ax1.get_xticks())
ax1.set_xticklabels([str(int(x)) for x in list(np.linspace(0, 255, cols_number)[::(cols_number // ticks)])])
ax1.set_title(f'Histogram with {cols_number} columns')
dft_matrix = self.get_dft(dft_mat_size)
ax2.imshow(dft_matrix, cmap='gray')
ax2.set_title(f'DFT with {dft_mat_size} matrix size')
dct_matrix = self.get_dct(dct_mat_size)
ax3.imshow(dct_matrix, cmap='gray')
ax3.set_title(f'DCT with {dct_mat_size} matrix size')
gradient = self.get_gradient(window_height)
ax4.plot(range(1, len(gradient)+1), gradient)
ax4.set_xlabel('window position')
ax4.set_ylabel('difference')
ax4.set_title(f'Gradient with {window_height} window height')
scaled_matrix = self.get_scale(scale)
ax5.imshow(scaled_matrix, cmap='gray')
ax5.set_title(f'Scale at {scale}% of original size')
_.axis('off')
if self.label_number is None:
title = f'Image №{self.order_number}'
else:
title = f'Image №{self.order_number} of person {self.label_number}'
fig.suptitle(title, fontsize=20)
plt.show()
def extract_features(image, method, param=None):
if not isinstance(image, Image):
print('Invalid arguments')
return
features = eval(f'image.{method}')
if param is None:
return features()
else:
return features(param)
def get_distance(feature_1, feature_2):
return cv2.norm(feature_1, feature_2)
orl_dir = os.path.join('.', 'ORL Face Database')
if __name__ == "__main__":
img_path_1 = os.path.join(orl_dir, 's23', '4.pgm')
# example_hist, _ = get_histogram(img)
'''
print(get_dct(img))
print(get_dft(img))
print(get_histogram(img))
print(get_gradient(img))
print(get_scale(img))
'''
img_path_2 = os.path.join(orl_dir, 's24', '5.pgm')
img_1 = Image(img_path_1, test=True)
img_2 = Image(img_path_2, test=True)
#print(img_1.__dict__['get_histogram'])
print(get_distance(img_1, img_2, 'get_histogram', 30))
img_1.show_features()
#plt.imshow(get_scale(img), cmap='gray')
#plt.show()