-
-
Notifications
You must be signed in to change notification settings - Fork 466
Expand file tree
/
Copy pathupload_dlib.py
More file actions
374 lines (319 loc) · 12.9 KB
/
Copy pathupload_dlib.py
File metadata and controls
374 lines (319 loc) · 12.9 KB
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
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
# coding=utf-8
import tensorflow as tf
import os, json, time
import numpy as np
from scipy import misc
from flask import Flask, request, url_for, make_response, abort, Response, jsonify, send_from_directory, redirect
from flask_sqlalchemy import SQLAlchemy
from werkzeug.utils import secure_filename
import facenet
import align.detect_face
import FaceProcessing
from save2gst import save2gst
from utilslib.aliyunUpload import aliyunUploadInit
from utilslib.aliyunUpload import aliyun_upload_img
# For Dlib Alignment
import cv2
from align import align_dlib
BASEDIR = os.getenv('RUNTIME_BASEDIR',os.path.abspath(os.path.dirname(__file__)))
UPLOAD_FOLDER = os.path.join(BASEDIR, 'image')
DATABASE = 'sqlite:///' + os.path.join(BASEDIR, 'data.sqlite')
ALLOWED_EXTENSIONS = set(['png', 'jpg', 'jpeg', 'gif', 'bitmap'])
app = Flask(__name__)
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
app.config['SQLALCHEMY_DATABASE_URI'] = DATABASE
app.config['SQLALCHEMY_COMMIT_ON_TEARDOWN'] = True
app.config['SQLALCHEMY_TRACK_MODIFICATIONS'] = True
db = SQLAlchemy(app)
image_size = 160
margin = 2
minsize = 20 # minimum size of face
threshold = [0.6, 0.7, 0.7] # three steps's threshold
factor = 0.709 # scale factor
confident_value = 0.78
USE_MTCNN = True
USE_DLIB = False
sess, graph = FaceProcessing.InitialFaceProcessor()
if USE_MTCNN:
graph2 = tf.Graph()
with graph2.as_default():
sess2 = tf.Session(config=tf.ConfigProto(log_device_placement=False), graph=graph2)
with sess2.as_default():
pnet, rnet, onet = align.detect_face.create_mtcnn(sess2, None)
if USE_DLIB:
dlibFacePredictor = os.path.join(BASEDIR,'../models',
"shape_predictor_68_face_landmarks.dat")
dlibAlign = align_dlib.AlignDlib(dlibFacePredictor)
def load_align_image(image_path, sess, graph, pnet, rnet, onet):
img = misc.imread(os.path.expanduser(image_path))
img_size = np.asarray(img.shape)[0:2]
with graph2.as_default():
# sess2 = tf.Session(config=tf.ConfigProto(log_device_placement=False))
with sess2.as_default():
bounding_boxes, _ = align.detect_face.detect_face(img, minsize, pnet, rnet, onet, threshold, factor)
nrof_faces = bounding_boxes.shape[0]
width = img_size[1]
height = img_size[0]
if nrof_faces > 0:
det = np.squeeze(bounding_boxes[0, 0:4])
bb = np.zeros(4, dtype=np.int32)
bb[0] = np.maximum(det[0] - margin / 2, 0)
bb[1] = np.maximum(det[1] - margin / 2, 0)
bb[2] = np.minimum(det[2] + margin / 2, img_size[1])
bb[3] = np.minimum(det[3] + margin / 2, img_size[0])
if bb[0] == 0 or bb[1] == 0 or bb[2] >= width or bb[3] >= height:
print('Out of boundary')
return None
print(bb)
print(img_size)
if USE_DLIB:
rgbImg = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
aligned = dlibAlign.align(160, rgbImg, bb)
else:
cropped = img[bb[1]:bb[3], bb[0]:bb[2], :]
aligned = misc.imresize(cropped, (image_size, image_size), interp='bilinear')
prewhitened = facenet.prewhiten(aligned)
return prewhitened
return None
def dlibImageProcessor(imgPath):
bgrImg = cv2.imread(imgPath)
if bgrImg is None:
raise Exception("Unable to load image: {}".format(imgPath))
rgbImg = cv2.cvtColor(bgrImg, cv2.COLOR_BGR2RGB)
# assert np.isclose(norm(rgbImg), 11.1355)
bb = dlibAlign.getLargestFaceBoundingBox(rgbImg)
#print(bb)
if bb is not None:
alignedFace = dlibAlign.align(160, rgbImg, bb)
prewhitened = facenet.prewhiten(alignedFace)
#plt.imshow(alignedFace)
#plt.show()
return alignedFace
return None
def featureCalculation(imgpath):
if USE_MTCNN:
img_data = load_align_image(imgpath, sess, graph, pnet, rnet, onet)
if img_data is not None:
with graph.as_default():
with sess.as_default():
embedding = FaceProcessing.FaceProcessingImageData(img_data, sess, graph)[0]
return embedding
if USE_DLIB:
img_data = dlibImageProcessor(imgpath)
if img_data is not None:
with graph.as_default():
with sess.as_default():
embedding = FaceProcessing.FaceProcessingImageData(img_data, sess, graph)[0]
return embedding
class People(db.Model):
__tablename__ = 'people'
id = db.Column(db.Integer, primary_key=True)
uuid = db.Column(db.String(64))
name = db.Column(db.String(64))
embed = db.Column(db.PickleType) # 存储任何Python对象，自动序列化
# filename = db.Column(db.String(64))
local_url = db.Column(db.String(128))
aliyun_url = db.Column(db.String(128))
def __repr__(self):
return '<People {}>'.format(self.id)
def updatePeopleImgURL(ownerid, url):
if len(url) < 1:
return
print(ownerid)
print(url)
with app.app_context():
man = People.query.filter_by(id=ownerid).first()
man.aliyun_url = url
db.session.add(man)
db.session.commit()
# db.session.query(People).filter(People.id == id).update({People.aliyun_url: url})
# db.session.commit()
"""
def dumps_embed(data):
return pickle.dumps(data)
def loads_embed(str):
return pickle.loads(str)
def hash_embed(data):
md5 = hashlib.md5()
md5.update(data)
return md5.hexdigest()
"""
def compare(emb1, emb2):
dist = np.sqrt(np.sum(np.square(np.subtract(emb1, emb2))))
# d = emb1 - emb2
# sqL2 = np.dot(d, d)
# print("+ Squared l2 distance between representations: {:0.3f}, dist is {:0.3f}".format(sqL2,dist))
print("+ distance between representations: {:0.3f}".format(dist))
# return sqL2
return dist
def allowed_file(filename):
"""
检查文件扩展名是否合法
:param filename:
:return: 合法 为 True
"""
return '.' in filename and \
filename.rsplit('.', 1)[1] in ALLOWED_EXTENSIONS
@app.route('/api/images/<filename>', methods=['GET'])
def img(filename):
# p = People.query.filter_by(filename=filename).first()
# if p and p.aliyun_url:
# return redirect(p.aliyun_url)
if os.path.isfile(os.path.join(app.config['UPLOAD_FOLDER'], filename)):
# 返回图片
return send_from_directory(app.config['UPLOAD_FOLDER'],
filename)
# 返回json
# data = {'img_name': filename, 'img_url': request.url}
# js = json.dumps(data)
# resp = Response(js, status=200, mimetype='application/json')
# return resp
else:
return abort(404)
@app.route('/api/images/', methods=['POST'])
def upload_img():
f = request.files['file'] # 从表单的file字段获取文件，file为该表单的name值
if f and allowed_file(f.filename): # 检查扩展名合法
filename = secure_filename(f.filename) # 修改成安全文件名
ext = filename.rsplit('.', 1)[1] # 获取文件后缀
unix_time = int(time.time())
uuid = request.args.get('uuid', '')
new_filename = uuid + str(unix_time) + '.' + ext # 修改了上传的文件名
imagepath = os.path.join(app.config['UPLOAD_FOLDER'], new_filename)
f.save(imagepath)
local_url = url_for('img', filename=new_filename, _external=True)
with graph.as_default():
# sess2 = tf.Session(config=tf.ConfigProto(log_device_placement=False))
with sess.as_default():
embedding = FaceProcessing.FaceProcessingOne(imagepath, sess, graph)[0]
# dumps_data = dumps_embed(embedding) # json序列化
people = People.query.all() # 数据库中所有的行
for p in people:
if compare(embedding, p.embed) < confident_value: # 比较, 小于confident_value认为同一个人
people = p
print(">>> same people: %d" % p.id)
break
else:
people = People(embed=embedding, uuid=uuid)
db.session.add(people)
db.session.commit()
print(">>> new people")
p_id = people.id # 识别出用户ID
people.aliyun_url = uploadImg.aliyun_upload_img(imagepath, 4)
# uploadImg.newjob(id, imagepath) # 异步线程上传图片到aliyun
people.local_url = local_url # 为ID对应的用户更新刚上传的图片
data = {'img_name': new_filename,
'local_url': local_url,
'aliyun_url': people.aliyun_url,
'errorn': 0,
'id': p_id,
'uuid': uuid,
} # 上传成功
js = json.dumps(data)
resp = Response(js, status=200, mimetype='application/json')
if people.aliyun_url: # 优先发送aliyun的图片地址
url = people.aliyun_url
else:
url = people.local_url
save2gst(uuid, id, url) # 调用故事帖api保存
return resp
else:
return jsonify({"errno": 1001, "message": "上传失败"})
@app.route('/api/fullimg/', methods=['POST'])
def upload_full_img():
f = request.files['file']
if f and allowed_file(f.filename):
ext = f.filename.rsplit('.', 1)[1]
unix_time = int(time.time())
uuid = request.args.get('uuid', '')
filename = uuid + str(unix_time) + '.' + ext
image_path = os.path.join(app.config['UPLOAD_FOLDER'], filename)
f.save(image_path)
if USE_MTCNN:
img_data = load_align_image(image_path, sess, graph, pnet, rnet, onet)
if USE_DLIB:
img_data = dlibImageProcessor(image_path)
if img_data is not None:
# print("Image Shape is ",img_data.shape)
with graph.as_default():
with sess.as_default():
embedding = FaceProcessing.FaceProcessingImageData(img_data, sess, graph)[0]
# print('Embedding of Full Image is %s' %embedding)
# 同步阻塞
people = People.query.all()
if people:
min_value = min([(compare(embedding, p.embed), p.id) for p in people]) # 遍历数据库并求最小compare值
print(min_value)
if min_value[0] < confident_value:
people = People.query.filter_by(id=min_value[1]).first()
print(">>> same people: %d" % min_value[1])
else:
people = People(embed=embedding, uuid=uuid)
db.session.add(people)
db.session.commit()
print(">>> new people")
else:
people = People(embed=embedding, uuid=uuid)
db.session.add(people)
db.session.commit()
print(">>> new people")
# people.filename = filename
local_url = url_for('img', filename=filename, _external=True)
people.local_url = local_url
aliyun_url = aliyun_upload_img(image_path, 4)
people.aliyun_url = aliyun_url
db.session.add(people)
db.session.commit()
#if len(aliyun_url) > 1: # 优先发送aliyun的图片地址
#save2gst(uuid, people.id, aliyun_url) # 发送请求给workai
data = {'error': 0,
'img_name': filename,
'id': people.id,
'uuid': people.uuid,
'aliyun_url': aliyun_url,
}
resp = Response(json.dumps(data), status=200, mimetype='application/json')
print('-----------------')
return resp
else:
print 'No available face in this image'
print '-----------------------'
os.remove(image_path)
return jsonify({"error": 1001, "message": u"识别失败"})
else:
return jsonify({'error': 1001, 'message': u'上传失败'})
@app.errorhandler(404)
def not_found(error=None):
message = {
'status': 404,
'message': 'Not Found ' + request.url,
}
return make_response(json.dumps(message), 404)
# 测试上传
@app.route('/test/upload')
def upload_test():
return '''
<!doctype html>
<title>Upload new File</title>
<h1>Upload new File</h1>
<form method=post action=/api/images enctype=multipart/form-data>
<p><input type=file name=file>
<input type=submit value=Upload>
</form>
'''
# flask默认的启动
if __name__ == '__main__':
uploadImg = aliyunUploadInit(updatePeopleImgURL)
if not os.path.exists(UPLOAD_FOLDER):
os.makedirs(UPLOAD_FOLDER)
if not os.path.exists(os.path.join(BASEDIR, 'data.sqlite')):
db.create_all()
# do nothing, just warm up
featureCalculation('./image/Mike_Alden_0001.png')
app.run(host='0.0.0.0',port=2999)
# gunicorn启动时创建文件
# if not os.path.exists(UPLOAD_FOLDER):
# os.makedirs(UPLOAD_FOLDER)
#
# if not os.path.exists(os.path.join(BASEDIR, 'data.sqlite')):
# db.create_all()