84648488
 Chunk
reverted.
|
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
|
__author__ = 'chunk'
from mdata import *
from mfeat import *
import os, sys
from PIL import Image
from hashlib import md5
import csv
import shutil
from common import *
import json
import collections
import happybase
class DataCV(DataDumperBase):
def __init__(self, base_dir='/home/hadoop/data/HeadShoulder/', sub_dir='Train/'):
DataDumperBase.__init__(self)
self.base_dir = base_dir
self.sub_dir = sub_dir
self.dict_data = {}
def format(self):
self.extract()
def _hash_copy(self, image, ispos):
if not image.endswith('jpg'):
img = Image.open(image)
img.save('res/tmp.jpg', format='JPEG')
image = 'res/tmp.jpg'
with open(image, 'rb') as f:
index = md5(f.read()).hexdigest()
self.dict_data[index] = ispos
# origion:
# dir = base_dir + 'Img/Train/' + index[:3]
dir = self.base_dir + self.sub_dir + 'Img/' + index[:3]
if not os.path.exists(dir):
os.makedirs(dir)
path = dir + '/' + index[3:] + '.jpg'
print path
if not os.path.exists(path):
shutil.copy(image, path)
else:
pass
def extract(self):
for path, subdirs, files in os.walk(self.base_dir + 'Orig/'):
for name in files:
imagepath = os.path.join(path, name)
print imagepath
if imagepath.split('/')[-2].startswith('pos'):
self._hash_copy(imagepath, True)
else:
self._hash_copy(imagepath, False)
ordict_img = collections.OrderedDict(sorted(self.dict_data.items(), key=lambda d: d[0]))
lstfile = self.base_dir + self.sub_dir + 'Img/Image.tsv'
with open(lstfile, 'w') as f:
tsvfile = csv.writer(f, delimiter='\t')
for key, value in ordict_img.items():
tsvfile.writerow([key] + [value])
def get_table(self, tablename, connection=None):
if connection is not None:
c = connection
else:
c = happybase.Connection('HPC-server')
tables = c.tables()
if tablename not in tables:
families = {'cf_pic': dict(),
'cf_info': dict(max_versions=10),
'cf_tag': dict(),
'cf_feat': dict(),
}
c.create_table(name=tablename, families=families)
tb = c.table(name=tablename)
return tb
def store_image(self, table):
timer.mark()
dir = self.base_dir + self.sub_dir + 'Img/'
maplst = dir + 'Image.tsv'
dict_databuf = {}
with open(maplst, 'rb') as tsvfile:
tsvfile = csv.reader(tsvfile, delimiter='\t')
for line in tsvfile:
path_img = self.base_dir + self.sub_dir + 'Img/Train/' + line[0][:3] + '/' + line[0][3:] + '.jpg'
if path_img:
with open(path_img, 'rb') as fpic:
dict_databuf[line[0] + '.jpg'] = fpic.read()
timer.report() # 58.761801s
timer.mark()
try:
with table.batch(batch_size=5000) as b:
for imgname, imgdata in dict_databuf.items():
b.put(imgname, {'cf_pic:data': imgdata})
raise ValueError("Something went wrong!")
except ValueError:
pass
timer.report() # 15.570524s
def store_tag(self, table):
timer.mark()
dir = self.base_dir + self.sub_dir + 'Img/'
maplst = dir + 'Image.tsv'
dict_tagbuf = {}
with open(maplst, 'rb') as tsvfile:
tsvfile = csv.reader(tsvfile, delimiter='\t')
for line in tsvfile:
dict_tagbuf[line[0] + '.jpg'] = line[1]
timer.report() # 0.009741s
timer.mark()
try:
with table.batch(batch_size=5000) as b:
for imgname, imgtag in dict_tagbuf.items():
b.put(imgname, {'cf_tag:class': imgtag})
raise ValueError("Something went wrong!")
except ValueError:
pass
timer.report() # 0.509696s
def get_feat(self, category='hog'):
dir = self.base_dir + self.sub_dir + 'Img/'
maplst = dir + 'images_map_Train.tsv'
dict_tagbuf = {}
with open(maplst, 'rb') as tsvfile:
tsvfile = csv.reader(tsvfile, delimiter='\t')
for line in tsvfile:
dict_tagbuf[line[0] + '.jpg'] = line[1]
dict_featbuf = {}
timer.mark()
for imgname, imgtag in dict_tagbuf.items():
# if imgtag == 'True':
path_img = self.base_dir + self.sub_dir + 'Img/Train/' + imgname[:3] + '/' + imgname[3:]
desc = FeatHOG.feat(path_img, size=(48, 48))
dict_featbuf[imgname] = desc
timer.report() # 4.337425s
timer.mark()
for imgname, desc in dict_featbuf.items():
# print imgname, desc
dir = self.base_dir + self.sub_dir + 'Feat/Train/' + imgname[:3] + '/'
if not os.path.exists(dir):
os.makedirs(dir)
featpath = dir + imgname[3:].split('.')[0] + '.' + category
with open(featpath, 'wb') as featfile:
featfile.write(json.dumps(desc.tolist()))
timer.report() # 14.862485s
def store_feat(self, table):
timer.mark()
dir = self.base_dir + self.sub_dir + 'Feat/'
dict_featbuf = {}
for path, subdirs, files in os.walk(dir + 'Train/'):
for name in files:
featpath = os.path.join(path, name)
# print featpath
with open(featpath, 'rb') as featfile:
imgname = path.split('/')[-1] + name.replace('.hog', '.jpg')
dict_featbuf[imgname] = featfile.read()
timer.report() # 0.577940s
timer.mark()
try:
with table.batch(batch_size=5000) as b:
for imgname, featdesc in dict_featbuf.items():
b.put(imgname, {'cf_feat:hog': featdesc})
raise ValueError("Something went wrong!")
except ValueError:
pass
timer.report() # 76.075477s
|