/
featureextractor.py
executable file
·256 lines (213 loc) · 8.52 KB
/
featureextractor.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
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
import cv2
from PIL import Image
import numpy
import skimage
from skimage.feature import daisy, hog
import subprocess
import overfeat
import numpy
from scipy.ndimage import imread
from scipy.misc import imresize
import database_file
import time
from find_obj import init_feature
from asift import affine_detect
print "Initializing overfeat"
overfeat.init('overfeat/overfeat/data/default/net_weight_0', 0)
#overfeat_initialized = False
def extract_overfeat(image_path):
# if overfeat_initialized == None or not overfeat_initialized:
# overfeat_initialized = True
print "Overfeat: ", image_path
image = imread(image_path)
print "Image shape: ", image.shape
if len(image.shape) == 2 or image.shape[2] == 2:
image = skimage.color.gray2rgb(image)
elif image.shape[2] == 4:
image_rgb = numpy.zeros((image.shape[0],image.shape[1], 3), numpy.uint8)
image_rgb[:,:,0] = image[:,:,0]
image_rgb[:,:,1] = image[:,:,1]
image_rgb[:,:,2] = image[:,:,2]
image = image_rgb
h0 = image.shape[0]
w0 = image.shape[1]
d0 = float(min(h0, w0))
image = image[int(round((h0-d0)/2.)):int(round((h0-d0)/2.)+d0),
int(round((w0-d0)/2.)):int(round((w0-d0)/2.)+d0), :]
image = imresize(image, (231, 231)).astype(numpy.float32)
#image = cv2.resize(image, (231, 231)).astype(numpy.float32)
# numpy loads image with colors as last dimension, transpose tensor
h = image.shape[0]
w = image.shape[1]
c = image.shape[2]
image = image.reshape(w*h, c)
image = image.transpose()
image = image.reshape(c, h, w)
print "Image size :", image.shape
out_categories = overfeat.fprop(image)
#layer 21,22,23
layer_output = overfeat.get_output(20)
print "Layer size: ", layer_output.shape
layer_output = layer_output.flatten()
descriptors = []
descriptors.append(layer_output)
out_categories = out_categories.flatten()
top = [(out_categories[i], i) for i in xrange(len(out_categories))]
top.sort()
print "\nTop classes :"
for i in xrange(5):
print(overfeat.get_class_name(top[-(i+1)][1]))
return descriptors
def extract_descriptor(image_path, thumbnail_size, feature_type="SIFT", descriptor_type="SIFT"):
t_start_cached = time.time()
descriptors_db = database_file.load_image(image_path, feature_type, descriptor_type, thumbnail_size)
t_end_cached = time.time() - t_start_cached
if descriptors_db == None:
print "Precomputing: " + str(image_path)
t_start = time.time()
features, descriptors = extract_descriptor_compute(image_path, thumbnail_size, feature_type, descriptor_type)
database_file.save_image(image_path, feature_type, descriptor_type, thumbnail_size, descriptors)
t_end = time.time() - t_start
print "Desc time: " + str(t_end)
return None, descriptors
else:
print "Cached: " + str(image_path)
print "Desc cached: " + str(t_end_cached)
return None, descriptors_db
def extract_descriptor_compute(image_path, thumbnail_size, feature_type="SIFT", descriptor_type="SIFT"):
if feature_type == "OVERFEAT" or descriptor_type == "OVERFEAT":
descriptor = extract_overfeat(image_path)
return None, descriptor
img = Image.open(image_path)
img.load()
if img == None:
raise Exception("Error: Cannot read " + image_path)
if img.size > thumbnail_size:
img.thumbnail(thumbnail_size)
img_array = numpy.array(img)
return extract_descriptor_process(img_array, thumbnail_size, feature_type, descriptor_type)
def extract_descriptor_process(image_array, thumbnail_size, feature_type="SIFT", descriptor_type="SIFT"):
#Converte para cinza se tiver mais que 3 canais (RGB)
img_gray = image_array
if len(image_array.shape) >= 3:
img_gray = cv2.cvtColor(image_array, cv2.COLOR_RGB2GRAY)
img_gray = numpy.array(img_gray, numpy.uint8)
if descriptor_type == "DAISY":
try:
daisy_descs = []
descs = daisy(img_gray, step=15, radius=15, rings=14, histograms=6, orientations=8)
for row in range(descs.shape[0]):
for col in range(descs.shape[1]):
daisy_descs.append(descs[row][col])
#print len(daisy_descs)
return None, daisy_descs
except:
print "Error while extracting Daisy descriptor"
return None, None
if descriptor_type == "HOG":
descs = hog(img_gray, orientations=8, pixels_per_cell=(16, 16), cells_per_block=(1, 1))
if len(descs) < 512:
return None, None
hog_descs = []
hog_descs.append(descs[0:648])
return None, hog_descs
if feature_type == "ASIFT" or descriptor_type == "ASIFT":
detector, matcher = init_feature('sift-flann')
kp, descs = affine_detect(detector, img_gray)
return kp, descs
#Normalize
#img_gray = img_gray.astype(numpy.float32)
#img_gray = (img_gray - img_gray.mean())/img_gray.std()
#img_gray = cv2.normalize(img_gray, img_gray,0, 255, cv2.NORM_MINMAX)
#img_gray = img_gray.astype(numpy.uint8)
featureDetector = cv2.FeatureDetector_create(feature_type)
descriptorDetector = cv2.DescriptorExtractor_create(descriptor_type)
keypoints = featureDetector.detect(img_gray)
keypoints, descriptors = descriptorDetector.compute(img_gray, keypoints)
return keypoints, descriptors
def extract_features_spatial(image_path, thumbnail_size, feature_type="SURF", descriptor_type="SURF"):
img = Image.open(image_path)
img.load()
if img == None:
raise Exception("Error: Cannot read " + image_path)
if img.size > thumbnail_size:
img.thumbnail(thumbnail_size)
img_array = numpy.array(img)
#Converte para cinza se tiver mais que 3 canais (RGB)
img_gray = img_array
if len(img_array.shape) >= 3:
img_gray = cv2.cvtColor(img_array, cv2.COLOR_RGB2GRAY)
img_gray = numpy.array(img_gray, numpy.uint8)
p1 = img_gray[0:img_gray.shape[0]/2, 0:img_gray.shape[1]/2]
p2 = img_gray[0:img_gray.shape[0]/2, img_gray.shape[1]/2:]
p3 = img_gray[img_gray.shape[0]/2:0, 0:img_gray.shape[1]/2]
p4 = img_gray[img_gray.shape[0]/2:0, img_gray.shape[1]/2:0]
patches = [p1, p2, p3, p4]
featureDetector = cv2.FeatureDetector_create(feature_type)
descriptorDetector = cv2.DescriptorExtractor_create(descriptor_type)
patches_keypoints = []
patches_descriptors = []
for patch in patches:
keypoints = featureDetector.detect(patch)
keypoints, descriptors = descriptorDetector.compute(patch, keypoints)
patches_keypoints.append(keypoints)
patches_descriptors.append(descriptors)
return patches_keypoints, patches_descriptors
def extract_ref2(image_path, thumbnail_size, num_subimages):
img = Image.open(image_path)
img.load()
if img == None:
raise Exception("Error: Cannot read " + image_path)
if img.size > thumbnail_size:
img.thumbnail(thumbnail_size)
img_array = numpy.array(img)
img_gray = img_array
if len(img_array.shape) >= 3:
img_gray = cv2.cvtColor(img_array, cv2.COLOR_RGB2GRAY)
img_gray = numpy.array(img_gray, numpy.uint8)
p1 = img_gray[0:img_gray.shape[0]/2, 0:img_gray.shape[1]/2]
p2 = img_gray[0:img_gray.shape[0]/2, img_gray.shape[1]/2:]
p3 = img_gray[img_gray.shape[0]/2:0, 0:img_gray.shape[1]/2]
p4 = img_gray[img_gray.shape[0]/2:0, img_gray.shape[1]/2:0]
patches = [p1, p2, p3, p4]
patch_total_hist = []
for patch in patches:
featureDetector = cv2.FeatureDetector_create("SURF")
keypoints = featureDetector.detect(img_gray)
kp_angles = []
if keypoints == None:
patch_hist = numpy.zeros((35),numpy.float32)
print "Patch hist: ", len(patch_hist)
for hist_item in patch_hist:
patch_total_hist.append(hist_item)
continue
for keypoint in keypoints:
kp_angles.append(keypoint.angle)
bins = range(0, 360, 10)
patch_hist = numpy.histogram(kp_angles, bins=bins, density=True)[0]
for hist_item in patch_hist:
patch_total_hist.append(hist_item)
patch_total_hist = numpy.asarray(patch_total_hist, numpy.float32)
return patch_total_hist
def extract_ref(image_path, thumbnail_size):
img = Image.open(image_path)
img.load()
if img == None:
raise Exception("Error: Cannot read " + image_path)
if img.size > thumbnail_size:
img.thumbnail(thumbnail_size)
img_array = numpy.array(img)
img_gray = img_array
if len(img_array.shape) >= 3:
img_gray = cv2.cvtColor(img_array, cv2.COLOR_RGB2GRAY)
img_gray = numpy.array(img_gray, numpy.uint8)
featureDetector = cv2.FeatureDetector_create("SURF")
keypoints = featureDetector.detect(img_gray)
kp_angles = []
if keypoints == None:
return []
for keypoint in keypoints:
kp_angles.append(keypoint.angle)
bins = range(0, 360, 10)
histogram = numpy.histogram(kp_angles, bins=bins, density=True)[0]
return histogram