-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathUi.py
278 lines (245 loc) · 11 KB
/
Ui.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
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
import sys
from PyQt5.QtWidgets import (QWidget, QToolTip, QPushButton, QApplication,QMessageBox)
from PyQt5.QtGui import QFont
from PyQt5.QtWidgets import QFileDialog
from PyQt5.QtCore import pyqtSlot
from PyQt5.Qt import QLineEdit
import os
from sklearn import model_selection
from sklearn import metrics
from sklearn.externals import joblib
import cv2
from scipy import misc
import tensorflow as tf
import numpy as np
import sys
import os
import copy
import argparse
import facenet
import align.detect_face
import restore
pathf = './haarcascade_frontalface_default.xml'
model_path ="./new_model"
face_cascade = cv2.CascadeClassifier(pathf)
class Example(QWidget):
def __init__(self):
super().__init__()
self.initUI()
def initUI(self):
QToolTip.setFont(QFont('SansSerif', 10))
self.setToolTip("This is a <b>QWidget</b> widget")
btn_add = QPushButton("Add a Person", self)
btn_add.setToolTip("Add a Person")
btn_add.resize(btn_add.sizeHint())
btn_add.move(50, 120)
btn_identify = QPushButton("Entrance Guard", self)
btn_identify.setToolTip("Entrance Guard")
btn_identify.resize(btn_identify.sizeHint())
btn_identify.move(50, 400)
btn_sample = QPushButton("Sample", self)
btn_sample.setToolTip("Sample the face")
btn_sample.resize(btn_sample.sizeHint())
btn_sample.move(50, 260)
self.setGeometry(800, 800, 800, 600)
self.setWindowTitle("Entrance Guard System")
self.show()
with tf.Graph().as_default():
with tf.Session().as_default() as sess:
# Load the model
facenet.load_model('model/20180408-102900')
images_placeholder = tf.get_default_graph().get_tensor_by_name("input:0")
embeddings = tf.get_default_graph().get_tensor_by_name("embeddings:0")
phase_train_placeholder = tf.get_default_graph().get_tensor_by_name("phase_train:0")
self.images_placeholder = images_placeholder
self.embeddings = embeddings
self.phase_train_placeholder = phase_train_placeholder
self.sess=sess
btn_add.clicked.connect(self.read)
btn_identify.clicked.connect(self.detect)
btn_sample.clicked.connect(self.sample)
def read(self):
file_name, ok = QFileDialog.getOpenFileNames(self, '多文件选择', './')
if ok:
if len(file_name)>9:
local_path = os.path.dirname(os.path.dirname(file_name[0]))
if os.path.exists(local_path+"/detect_face")!=0 :
if os.path.exists( local_path+"/embedding.txt")==0:
embedding_image(local_path,self.images_placeholder,self.embeddings,self.phase_train_placeholder,self.sess)
train_model(local_path)
if os.path.exists(local_path+"/detect_face")==0:
detect_face(local_path,file_name)
embedding_image(local_path,self.images_placeholder,self.embeddings,self.phase_train_placeholder,self.sess)
#for roots,dirs,files in os.walk("./new_model"):
#print(dirs)
#print(os.path.dirname(file_name[0]))
train_model(os.path.dirname(file_name[0]))
QMessageBox.information(self, "Message", "success",QMessageBox.Ok, QMessageBox.Ok)
def detect(self):
cap = cv2.VideoCapture(0)
frame_time = 0
print("开始识别")
while True:
ret, frame = cap.read()
timeF = 1
if (frame_time % timeF == 0):
print(11111111111111111111111111111111111111111)
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
if (len(faces) != 0):
images = image_align(faces,frame,160,44)
cv2.imshow('img', frame)
cv2.waitKey(1)
if frame_time>80:
feed_dict = {self.images_placeholder: images, self.phase_train_placeholder: False}
emb = self.sess.run(self.embeddings, feed_dict=feed_dict)
model_list = get_model_name(model_path)
access_flag = 0
for model_name in model_list:
model = joblib.load(model_name)
predict = model.predict(emb)
if (predict == 0):
print("Access")
access_flag = 1
break
if (predict == 1):
print("Denied")
continue
if access_flag ==0:
QMessageBox.information(self, "Message", "Denied", QMessageBox.Ok, QMessageBox.Ok)
break
else:
QMessageBox.information(self, "Message", "Access", QMessageBox.Ok, QMessageBox.Ok)
break
print("GGGGGGGGGGGGGGGGGGG")
frame_time = frame_time + 1
# 第一部分：抽帧（暂定一秒5帧）将检测的脸部存成image
# 第二部分：在将img输入facenet并获得embedding向量
# 第三部分：利用得到的embedding训练一个分类模型
# 第四部分：进行脸部识别并绘制
cap.release()
cv2.destroyAllWindows()
def sample(self):
name = input("please input the name!")
if os.path.exists(model_path+"/"+name)==0:
os.mkdir(model_path+"/"+name)
os.mkdir(model_path+"/"+name+"/detect_face")
cap = cv2.VideoCapture(0)
frame_count = 0
pic_count = 0
print("开始采样")
while True:
ret, frame = cap.read()
timeF = 1
#每10张采样一次
sample_interval = 10
if (frame_count % timeF == 0):
gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
if len(faces)!=0:
images = image_align(faces, frame, 160, 44)
# cv2.putText(frame, "please look at the cap", (50, 50),
# cv2.FONT_HERSHEY_COMPLEX_SMALL, 2, (255, 0, 0),
# thickness=1.5, lineType=2)
cv2.imshow('img', frame)
cv2.waitKey(1)
if pic_count ==40:
break
if frame_count%sample_interval == 0:
pic_count = pic_count+1
pic_path = model_path+"/"+name+"/detect_face"+"/"+str(pic_count)+".jpg"
for (x, y, w, h) in faces:
cv2.imwrite(pic_path,frame[y:y+h,x:x+w])
frame_count = frame_count +1
cap.release()
cv2.destroyAllWindows()
def embedding_image(path, images_placeholder, embeddings, phase_train_placeholder,sess):
if os.path.exists(path + "/embedding.txt") == 0:
path_detect = path + "/detect_face"
face_path = get_file_name(path_detect)
# print(face_path)
images_align = restore.image_resize_no_align(face_path, image_size=160, margin=44)
feed_dict = {images_placeholder: images_align, phase_train_placeholder: False}
emb = sess.run(embeddings, feed_dict=feed_dict)
print(emb)
np.savetxt(path + "/embedding.txt", emb)
def train_model(file_path):
other_path = os.path.dirname(file_path)+"/other/embedding.txt"
sample_path = file_path+"/embedding.txt"
print(other_path)
print(sample_path)
if os.path.exists(sample_path) != 0 :
X = []
f = open(sample_path)
for line in f:
list = line.strip('\n').split(' ')
X.append(list)
sample_length = len(X)
f1 = open(other_path)
for line in f1:
list_other = line.strip('\n').split(' ')
X.append(list_other)
other_length = len(X)-sample_length
train_y=np.zeros(len(X))
for i in range(sample_length):
train_y[i]=0
for i in range(other_length):
train_y[sample_length+i]=1
X_train, X_test, y_train, y_test = model_selection.train_test_split(X, train_y, test_size=.2, random_state=42)
# KNN Classifier
def knn_classifier(train_x, train_y):
from sklearn.neighbors import KNeighborsClassifier
model = KNeighborsClassifier()
model.fit(train_x, train_y)
return model
classifiers = knn_classifier
model = classifiers(X_train, y_train)
predict = model.predict(X_test)
accuracy = metrics.accuracy_score(y_test, predict)
print('accuracy: %.2f%%' % (100 * accuracy))
# 保存模型
joblib.dump(model, file_path+"/knn.model")
def detect_face(path,image_name):
if os.path.exists(path+"/detect_face")==0:
os.mkdir(path+"/detect_face")
face_num = 0
for file in image_name:
img = cv2.imread(file)
#print(file)
#cv2.imshow('img', img)
#cv2.waitKey(500)
gray = cv2.imread(file, 0)
faces = face_cascade.detectMultiScale(gray, 1.3, 5)
if len(faces) == 0:
continue
for (x, y, w, h) in faces:
cv2.imwrite(path+"/detect_face/" + str(face_num) + ".jpg", img[y:y + h, x:x + w])
face_num = face_num + 1
def get_file_name(file_dir):
L=[]
for root, dirs, files in os.walk(file_dir):
for file in files:
if os.path.splitext(file)[1] == '.jpg':
L.append(os.path.join(root, file))
return L
def get_model_name(file_dir):
L=[]
for root, dirs, files in os.walk(file_dir):
for file in files:
if os.path.splitext(file)[1] == '.model':
L.append(os.path.join(root, file))
return L
def image_align(faces,frame,image_size,margin):
img_list = []
for (x, y, w, h) in faces:
img = frame[y:y + h, x:x + w]
cv2.rectangle(frame, (x, y), (x + w, y + h), (255, 0, 0), 2)
aligned = misc.imresize(img, (image_size, image_size), interp='bilinear')
prewhitened = facenet.prewhiten(aligned)
img_list.append(prewhitened)
images = np.stack(img_list)
return images
if __name__ == "__main__":
app = QApplication(sys.argv)
ex = Example()
sys.exit(app.exec_())