-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmain.py
343 lines (225 loc) · 10.9 KB
/
main.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
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
import cv2
import mediapipe as mp
import numpy as np
import math
from time import time
class face_summetry():
def __init__(self):
# Face Mesh
mp_face_mesh = mp.solutions.face_mesh
self.face_mesh = mp_face_mesh.FaceMesh(static_image_mode = True)
self.image_captured = False
self.captured_image = ''
self.display_text = ''
self.start_time = time()
self.end_time = time()
self.timer_set = False
def draw_face_landmarks(self,img):
# Facial landmarks
results = self.face_mesh.process(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
if not results.multi_face_landmarks == None:
landmarks = results.multi_face_landmarks[0]
for idx, landmark in enumerate(landmarks.landmark):
if idx>0:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
img = cv2.circle(img, (relative_x,relative_y), 2, (0,255,0), -1)
return img
def get_final_result(self,img):
# Facial landmarks
results = self.face_mesh.process(cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
horizontal_pts = []
horizontal_pts_eyes = []
vertical_pts = []
left_lip_sym = []
right_lip_sym = []
left_eye_sym = []
right_eye_sym = []
if not results.multi_face_landmarks == None:
landmarks = results.multi_face_landmarks[0]
h,w,_ = img.shape
for idx, landmark in enumerate(landmarks.landmark):
if idx == 13:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
img = cv2.line(img, (relative_x,0), (relative_x,h), (0,0,0), 1)
for pts in range(0,h,5):
vertical_pts.append((relative_x,pts))
#Eyes
if idx == 246:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
img = cv2.line(img, (0,relative_y), (w,relative_y), (0,0,0), 1)
for pts in range(0,w,5):
horizontal_pts_eyes.append((pts,relative_y))
#Lips
if idx == 61:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
img = cv2.line(img, (0,relative_y), (w,relative_y), (0,0,0), 1)
for pts in range(0,w,5):
horizontal_pts.append((pts,relative_y))
for idx, landmark in enumerate(landmarks.landmark):
#Left lip
if idx == 308:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
for h_points in horizontal_pts:
left_lip_sym.append(math.hypot(h_points[0] - relative_x, h_points[1] - relative_y))
img = cv2.circle(img, (relative_x,relative_y), 2, (0,0,255), -1)
#Right lip
if idx == 61:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
for h_points in horizontal_pts:
right_lip_sym.append(math.hypot(h_points[0] - relative_x, h_points[1] - relative_y))
img = cv2.circle(img, (relative_x,relative_y), 2, (0,0,255), -1)
#Right eye
if idx == 246:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
for h_points in horizontal_pts_eyes:
right_eye_sym.append(math.hypot(h_points[0] - relative_x, h_points[1] - relative_y))
img = cv2.circle(img, (relative_x,relative_y), 2, (0,0,255), -1)
#Left eye
if idx == 466:
x = landmark.x
y = landmark.y
relative_x = int(img.shape[1] * x)
relative_y = int(img.shape[0] * y)
img = cv2.circle(img, (relative_x,relative_y), 2, (0,0,255), -1)
for h_points in horizontal_pts_eyes:
left_eye_sym.append(math.hypot(h_points[0] - relative_x, h_points[1] - relative_y))
count = 0
cv2.putText(img,'Asymmetries found:',(10,20),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
if min(left_lip_sym) >2:
count+=1
cv2.putText(img,'Left lip',(10,40),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
if min(right_lip_sym) >2:
if count == 0:
cv2.putText(img,'Right lip',(10,40),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
elif count == 1:
cv2.putText(img,'Right lip',(10,60),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
count+=1
if min(left_eye_sym) >2:
if count == 0:
cv2.putText(img,'Left eye',(10,40),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
elif count == 1:
cv2.putText(img,'Left eye',(10,60),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
elif count == 2:
cv2.putText(img,'Left eye',(10,80),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
count += 1
if min(right_eye_sym) >2:
if count == 0:
cv2.putText(img,'Right eye',(10,40),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
if count == 1:
cv2.putText(img,'Right eye',(10,60),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
if count == 2:
cv2.putText(img,'Right eye',(10,80),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
if count == 3:
cv2.putText(img,'Right eye',(10,100),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
#Change accuracy
if min(left_lip_sym) <=2 and min(right_eye_sym) <=2 and min(left_eye_sym)<=2 and min(right_lip_sym)<=2:
cv2.putText(img,'None',(10,40),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
return img
def head_tilt_detection(self,image):
image = cv2.cvtColor(cv2.flip(image, 1), cv2.COLOR_BGR2RGB)
# To improve performance
image.flags.writeable = False
# Get the result
results = self.face_mesh.process(image)
# To improve performance
image.flags.writeable = True
# Convert the color space from RGB to BGR
image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
img_h, img_w, _ = image.shape
face_3d = []
face_2d = []
if results.multi_face_landmarks:
for face_landmarks in results.multi_face_landmarks:
for idx, lm in enumerate(face_landmarks.landmark):
if idx == 33 or idx == 263 or idx == 1 or idx == 61 or idx == 291 or idx == 199:
x, y = int(lm.x * img_w), int(lm.y * img_h)
# Get the 2D Coordinates
face_2d.append([x, y])
# Get the 3D Coordinates
face_3d.append([x, y, lm.z])
# Convert it to the NumPy array
face_2d = np.array(face_2d, dtype=np.float64)
# Convert it to the NumPy array
face_3d = np.array(face_3d, dtype=np.float64)
# The camera matrix
focal_length = 1 * img_w
cam_matrix = np.array([ [focal_length, 0, img_h / 2],
[0, focal_length, img_w / 2],
[0, 0, 1]])
# The distortion parameters
dist_matrix = np.zeros((4, 1), dtype=np.float64)
# Solve PnP
_, rot_vec, _ = cv2.solvePnP(face_3d, face_2d, cam_matrix, dist_matrix)
# Get rotational matrix
rmat, _ = cv2.Rodrigues(rot_vec)
# Get angles
angles, _, _, _, _, _ = cv2.RQDecomp3x3(rmat)
# Get the y rotation degree
x = angles[0] * 360
y = angles[1] * 360
z = angles[2] * 360
# See where the user's head tilting
if x < -2 or x >2 or y>1 or y<-1:
print('Head tilted')
return True
return False
def midpoint(self,p1, p2):
return (p1[0]+p2[0])/2, (p1[1]+p2[1])/2
def main_loop(self):
cap = cv2.VideoCapture(0,cv2.CAP_DSHOW)
while True:
ret, img = cap.read()
final_img = img.copy()
if not ret:
print('=============Camera not found================')
break
if not self.head_tilt_detection(img):
img = self.draw_face_landmarks(img)
if not self.timer_set:
self.timer_set = True
self.start_time = time()
self.end_time = time()
if self.end_time - self.start_time >=5:
final_img = self.get_final_result(final_img)
cv2.imwrite('test.jpg',final_img)
cv2.destroyAllWindows()
cv2.imshow('Final output',final_img)
cv2.waitKey()
cv2.destroyAllWindows()
return
else:
self.start_time = time()
self.end_time = time()
cv2.putText(img,'Keep your head straight and smile',(10,20),cv2.FONT_HERSHEY_PLAIN,1,(255,0,0),1)
cv2.imshow('Live feed',img)
cv2.waitKey(10)
def main_loop_for_img(self,path):
img = cv2.imread(path)
# self.draw_face_landmarks(img)
img = self.get_final_result(img)
cv2.imshow('Live feed',img)
cv2.waitKey()
cv2.imwrite('output.jpg',img)
objectt = face_summetry()
objectt.main_loop()