-
Notifications
You must be signed in to change notification settings - Fork 0
/
budz_cnn_predictor6.py
315 lines (292 loc) · 12.8 KB
/
budz_cnn_predictor6.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
# USAGE
#frameworkpython3 ./budz_cnn_predictor6.py --shape-predictor ./budz_predictor_landmarks.dat --video ../../face_recognition-master/examples/bbunny-clip0.mp4 --output ./budz_predictor0.avi
# import the necessary packages
from imutils import face_utils
import numpy as np
import argparse
import imutils
import dlib
import cv2
import math
from random import randint
def vidinit(args):
stream = cv2.VideoCapture(args["video"])
length = int(stream.get(cv2.CAP_PROP_FRAME_COUNT))
writer = None
# Find OpenCV version
(major_ver, minor_ver, subminor_ver) = (cv2.__version__).split('.')
if int(major_ver) < 3 :
fps = stream.get(cv2.cv.CV_CAP_PROP_FPS)
print("Frames per second using video.get(cv2.cv.CV_CAP_PROP_FPS): {0}".format(fps))
else :
fps = stream.get(cv2.CAP_PROP_FPS)
print("Frames per second using video.get(cv2.CAP_PROP_FPS) : {0}".format(fps))
return stream, length, writer, fps
def frameincrement(frame_number):
(grabbed, frame) = stream.read()
frame_number += 1
print("grab frame, frame# = {}".format(frame_number))
# if not grabbed, then no more stream
return grabbed, frame, frame_number
def dispwrite(writer, args, frame, fps):
# if the video writer is None *AND* we are supposed to write
# the output video to disk initialize the writer
if writer is None and args["output"] is not None:
print("init writer")
fourcc = cv2.VideoWriter_fourcc(*"MJPG")
writer = cv2.VideoWriter(args["output"], fourcc, fps, (frame.shape[1], frame.shape[0]), True)
if args["display"] > 0:
cv2.imshow("Frame", frame)
# cv2.imshow("Frame", open_cv_image)
cv2.waitKey(int(1000/fps))
# key = cv2.waitKey(1) & 0xFF
# if the `q` key was pressed, break from the loop
# if key == ord("q"):
# break
if writer is not None:
writer.write(frame)
print("writing frame")
return frame, writer
def calculateDistance(x1,y1,x2,y2):
dist = math.sqrt((x2 - x1)**2 + (y2 - y1)**2)
return dist
def face_shaper(rgb_img, face):
shape = predictor(rgb_img, face)
embedding = recognizer.compute_face_descriptor(rgb_img, shape)
embedding = [x for x in embedding]
embedding = np.array(embedding, dtype="float32")[np.newaxis, :]
shape = face_utils.shape_to_np(shape)
# label = recognize_face(embedding, embeddings, labels)
return shape
# def drawfeat(shape, leyepos, reyepos, nosepos, frame):
def drawfeat(shape, frame):
e1cx = int((shape[0][0] - shape[1][0])/ 2 + shape[1][0])
e2cx = int((shape[2][0] - shape[3][0])/ 2 + shape[3][0])
ncy = int((shape[0][1] - shape[4][1])/ 2 + shape[2][1])
radius = (shape[0][0] - shape[1][0]) / 2
r1 = int(radius * 2)
r2 = int(radius * 1.5)
nr = int(radius * 3)
for e in range(1000):
e1x = randint(-35, 30) + e1cx
e1y = randint(-30, 35) + int(shape[1][1])
e1r = randint(100, 165)
e1g = randint(50, 120)
e1b = randint(0, 70)
e1pos = (e1x, e1y)
e1clr = (e1b, e1g, e1r)
leyepos.append(e1pos)
lcolor.append(e1clr)
e2x = randint(-25, 20) + e2cx
e2y = randint(-20, 20) + int(shape[3][1])
e2r = randint(100, 165)
e2g = randint(50, 120)
e2b = randint(0, 70)
e2pos = (e2x, e2y)
e2clr = (e2b, e2g, e2r)
reyepos.append(e2pos)
rcolor.append(e2clr)
nx = randint(-20, 20) + int(shape[4][0])
ny = randint(-10, 20) + int(shape[4][1])
nr = randint(100, 165)
ng = randint(50, 120)
nb = randint(0, 70)
npos = (nx, ny)
nclr = (nb, ng, nr)
nosepos.append(npos)
ncolor.append(nclr)
if len(nosepos) > 2000:
nosepos[:-1].copy()
ncolor[:-1].copy()
reyepos[:-1].copy()
rcolor[:-1].copy()
leyepos[:-1].copy()
lcolor[:-1].copy()
print("lcolor {}".format(lcolor[len(lcolor) - 1]))
for i, pos in enumerate(leyepos):
x1, y1 = list(pos)
# print("x1: {}, y1: {}".format(x1, y1))
# cv2.circle(frame, (x1, y1), 1, lcolor[i], -1)
if ((x1 - e1cx) ** 2) / (r2 ** 2) + ((y1 - int(shape[1][1])) ** 2) / (r1 ** 2) <= 1 :
# cv2.circle(frame, pos, randint(1, 5), lcolor[len(lcolor) - 1], -1)
cv2.circle(frame, pos, randint(1, 5), lcolor[randint(0, len(lcolor) - 1)], -1)
center = (e1cx, int(shape[1][1]))
if r1 >= 50:
r1 = int((shape[0][0] - shape[1][0])/1)
r2 = int(r1 / 1.5)
axes = (r1, r2)
rotation_angle = randint(0, 10)
start_angle = randint(0, 360)
end_angle = randint(30, 360)
color = (randint(30, 120), randint(50, 140), randint(100, 165))
thickness = 1
cv2.ellipse(frame, center, axes, rotation_angle, start_angle, end_angle, color, thickness)
while r1 < 50:
r1 += 1
r2 += 1
for i, pos in enumerate(reyepos):
x1, y1 = list(pos)
# print("x1: {}, y1: {}".format(x1, y1))
# cv2.circle(frame, (x1, y1), 1, lcolor[i], -1)
if ((x1 - e2cx) ** 2) / (r1 ** 2) + ((y1 - int(shape[1][1])) ** 2) / (r2 ** 2) <= 1 :
cv2.circle(frame, pos, randint(1, 5), rcolor[i], -1)
center = (e2cx, int(shape[1][1]))
if r1 >= 50:
r1 = int((shape[0][0] - shape[1][0])/1)
r2 = int(r1 / 1.5)
axes= (r1, r2)
rotation_angle = randint(0, 10)
start_angle = randint(0, 360)
end_angle = randint(30, 360)
color = (randint(30, 120), randint(50, 140), randint(100, 165))
thickness = 1
print("center {}, axes {}, rotation_angle {}, start_angle {}, end_angle {}, color {}, thickness {}" .format(center, axes, rotation_angle, start_angle, end_angle, color, thickness))
cv2.ellipse(frame, center, axes, rotation_angle, start_angle, end_angle, color, thickness)
while r1 < 50:
r1 += 1
r2 += 1
for i, pos in enumerate(nosepos):
x1, y1 = list(pos)
# print("x1: {}, y1: {}".format(x1, y1))
# cv2.circle(frame, (x1, y1), 1, lcolor[i], -1)
if ((x1 - nx) ** 2) / (r1 ** 2) + ((y1 - int(shape[1][1])) ** 2) / (r2 ** 2) <= 1 :
cv2.circle(frame, pos, randint(1, 5), ncolor[i], -1)
center = (int(shape[4][0]), int(shape[1][1]))
if r1 >= 50:
r1 = int((shape[0][0] - shape[1][0])/1)
r2 = int(r1 / 1.5)
axes= (r1, r2)
rotation_angle = randint(0, 10)
start_angle = randint(0, 360)
end_angle = randint(30, 360)
color = (randint(30, 120), randint(50, 140), randint(100, 165))
thickness = 1
cv2.ellipse(frame, center, axes, rotation_angle, start_angle, end_angle, color, thickness)
while r1 < 50:
r1 += 1
r2 += 1
# if ((xe1 - cxe1) ** 2) / (r1 ** 2) + ((ye1 - int(shape[1][1])) ** 2) / (r2 ** 2) <= 1 :
# cv2.circle(frame, (xe1, ye1), 1, (e1b, e1g, e1r), -1)
# if ((xe2 - cxe2) ** 2) / (r1 ** 2) + ((ye2 - int(shape[3][1])) ** 2) / (r2 ** 2) <= 1 :
# cv2.circle(frame, (xe2, ye2), 1, (e2b, e2g, e2r), -1)
# if ((xn - int(shape[4][0])) ** 2) / (r1 ** 2) + ((yn - int(shape[4][1])) ** 2) / (rn ** 2) <= 1 :
# cv2.circle(frame, (xn, yn), 1, (nb, ng, nr), -1)
# for (i, (x, y)) in enumerate(shape):
# if i % 5 == 0:
# for j in range(300):
# px = randint(-5, 5) + x
# py = randint(-5, 5) + y
# cv2.circle(frame, (px, py), 1, (255, 255, 255), -1)
# if i % 5 == 1:
# for j in range(300):
# px = randint(-5, 5) + x
# py = randint(-5, 5) + y
# cv2.circle(frame, (px, py), 1, (0, 0, 255), -1)
#
# if i % 5 == 4:
# for n in range(500):
# xn = randint(-100, 100) + x
# yn = randint(-100, 100) + y
# if (((xn - x) ** 2) / (r2 ** 2) + (yn - y) ** 2) / (r1 ** 2) <= 1 :
# cv2.circle(frame, (nosex, nosey), 1, (0, 0, 0), -1)
return frame
ap = argparse.ArgumentParser()
ap.add_argument("-p", "--shape-predictor", required=True,
help="path to facial landmark predictor")
ap.add_argument("-v", "--video", required=True,
help="path to input video")
ap.add_argument("-o", "--output", type=str,
help="path to output video")
ap.add_argument('-w', '--weights', default='./mmod_human_face_detector.dat',
help='path to weights file')
ap.add_argument("-d", "--display", type=int, default=1,
help="whether or not to display output frame to screen")
args = vars(ap.parse_args())
stream, length, writer, fps = vidinit(args)
# points = []
leyepos = []
reyepos = []
nosepos = []
lcolor = []
rcolor = []
ncolor = []
detector = dlib.cnn_face_detection_model_v1(args["weights"])
#hog
#detector = dlib.get_frontal_face_detector()
# detector = dlib.fhog_object_detector("bb0.svm")
# detector = dlib.simple_object_detector("b2detector.svm")
predictor = dlib.shape_predictor(args["shape_predictor"])
recognizer = dlib.face_recognition_model_v1("./dlib_face_recognition_resnet_model_v1.dat")
frame_number = 0
grabbed, frame, frame_number = frameincrement(frame_number)
print("video dims: {} x {}".format(frame.shape[0], frame.shape[1]))
# ratio = 1/(500/vid_w)
# print("1:1 ratio: {}".format(ratio))
while True:
if not grabbed:
break
# image = cv2.imread(args["image"])
# frame = imutils.resize(frame, width=500)
# gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
rgb_img = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
faces = detector(rgb_img, 1)
faces = [face.rect for face in faces]
if len(faces) == 0:
print("no face detected, frame to write: frame# {}".format(frame_number))
frame, writer = dispwrite(writer, args, frame, fps)
grabbed, frame, frame_number = frameincrement(frame_number)
else:
print("# of faces: {}" .format(len(faces)))
for face in faces:
# x,y,w,h = face.left(), face.top(), face.right(), face.bottom()
# cv2.rectangle(frame, (x,y), (w,h), (255,200,150), 2, cv2.LINE_AA)
(x1, y1, x2, y2) = face.left(), face.top(), face.right(), face.bottom()
r = int((x2 - x1)/2)
r1 = int(r * 1.2)
r2 = int(r * 1)
cx = int(x1 + r / .9)
cy = int(y1 + r / 1.8)
# d = int(r * 1.414)
center = (cx, cy)
for spiral in range(10):
r1 = r1 - spiral
r2 = r2 - spiral
axes= (r1, r2)
rotation_angle = randint(-30, 30)
start_angle = randint(0, 360)
end_angle = randint(30, 360)
color = (randint(20, 50), randint(50, 150), randint(150, 255))
thickness = 1
cv2.ellipse(frame, center, axes, rotation_angle, start_angle, end_angle, color, thickness)
shape = face_shaper(rgb_img, face)
frame = drawfeat(shape, frame)
# for p in range(100):
# shiftx = randint(-r1, r1) + cx
# shifty = randint(-r2, r2) + cy
# # dist = calculateDistance(cx, cy, shiftx, shifty)
# # if dist < r:
# # cv2.circle(frame, (shiftx, shifty), 2, (0, 0, 255), -1)
# shiftxy = (shiftx, shifty)
# if len(points) > 2000:
# del points[(len(points) - 1)]
# points.append(shiftxy)
# for point in points:
# print("point in face: {}" .format(point))
# if ((point[0] - cx) ** 2) / (r1 ** 2) + ((point[1] - cy) ** 2) / (r2 ** 2) <= 1 :
# thickness = randint(1, 3)
# red = randint(150, 255)
# green = randint(100, 150)
# blue = randint(50, 100)
# dots = cv2.circle(frame, (point[0], point[1]), thickness, (blue, green, red), -1)
# print("dots: {}".format(dots))
print("face scribbled with {} points, frame to write: frame# {}".format(len(nosepos), frame_number))
frame, writer = dispwrite(writer, args, frame, fps)
grabbed, frame, frame_number = frameincrement(frame_number)
# save output image
# cv2.imwrite("cnn_face_detection.png", image)
# # close all windows
# cv2.destroyAllWindows()
stream.release();
if writer is not None:
print("writer release")
writer.release()