-
Notifications
You must be signed in to change notification settings - Fork 0
/
master.py
211 lines (184 loc) · 7.37 KB
/
master.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
# coding: utf-8
import cv2
import numpy as np
import sys, os
import time, math
#list of expression names
# faces = ["surprised", "smile", "sad", "crying", "laugh", "smug"]
faces = ["surprised", "smile", "sad", "crying", "laugh", "smug", "nice", "wink"]
detected_counts = [0,0,0,0,0,0,0,0]
# A-KAZE検出器の生成
akaze = cv2.AKAZE_create()
# カメラ
cap = cv2.VideoCapture(int(sys.argv[1]))
cap_w = cap.get(3)
cap_h = cap.get(4)
# cap.set(3, cap_w)
# cap.set(4,cap_h)
# cap.set(5, 30)
#width and height for red box
window_width = int(sys.argv[2])
window_height = int(sys.argv[2])
img3 = None
now_str = None
# def display_face(ura_count,img2):
# global HSV
# ura_count = min([ura_count*0.1,1])
# HSV[:,:,1:] = HSV[:,:,1:]*(1-fltr*ura_count) + emo[:,:,1:]*fltr*ura_count
# img2 = cv2.cvtColor(HSV,cv2.COLOR_HSV2BGR)
# return img2
# function for taking a picture of our own expressions
def my_capture():
global now_str
cap = cv2.VideoCapture(int(sys.argv[1]))
cap_w = cap.get(3)
cap_h = cap.get(4)
# cap.set(3, cap_w)
# cap.set(4,cap_h)
# cap.set(5, 30)
#windowsize
window_width = int(sys.argv[2])
window_height = int(sys.argv[2])
i = 0
c_faces = ["smile", "sad", "cry", "wink", "angry"]
now = time.ctime()
cnvtime = time.strptime(now)
now_str = (time.strftime("%Y-%m-%d_%H-%M-%S", cnvtime))
os.mkdir("output/" + now_str)
flg = False
sec = 0
#時間間隔, 何秒ごとに撮影するか
gap = 5
while(cap.isOpened()):
ret, c_frame = cap.read()
cv2.rectangle(c_frame, (int(cap_w/2 - window_width/2), int(cap_h/2 - window_height/2)), (int(cap_w/2 + window_width/2), int(cap_h/2 + window_height/2)), (0,0,255), thickness = 2)
# flg to start shooting
if flg:
sec = math.floor((time.time() - start))
else:
cv2.putText(c_frame, "Press s to start.", (int(cap_w/4), int(cap_h/2)), cv2.FONT_HERSHEY_SIMPLEX, 3.0, (255, 255, 255), thickness=4)
cv2.putText(c_frame, c_faces[i], (100, 100), cv2.FONT_HERSHEY_SIMPLEX, 2.0, (255, 0, 0), thickness=2)
cv2.putText(c_frame, "Countdown :"+str((i+1)*5 - sec), (800, 100), cv2.FONT_HERSHEY_SIMPLEX, 2.0, (255, 255, 255), thickness=2)
# 画像表示
cv2.imshow('frame', c_frame)
key_pressed = cv2.waitKey(1)
# if key_pressed & 0xFF == ord('c'):
#5秒たったら
if sec == (i+1)*5:
cv2.imwrite("output/" + now_str + "/" + c_faces[i] + ".jpg", c_frame[int(cap_h/2 - window_height/2):int(cap_h/2 + window_height/2),int(cap_w/2 - window_width/2):int(cap_w/2 + window_width/2)])
# cv2.imwrite("/Volumes/斉藤のパブリックフォルダ/output/" + "happy" + "_" + str(int(time.time())).replace(".","") + ".jpg", c_frame[int(cap_h/2 - window_height/2):int(cap_h/2 + window_height/2),int(cap_w/2 - window_width/2):int(cap_w/2 + window_width/2)])
i += 1
if i == len(c_faces): break
if key_pressed & 0xFF == ord('q'):
break
elif key_pressed & 0xFF == ord("s"):
start = time.time()
flg = True
# cap.release()
# cv2.destroyAllWindows()
def display_face(ura_count, img2):
global now_str
if ura_count > 2:
global img3
img3 = cv2.resize(cv2.imread("output/" + now_str + "/" + "smile" + ".jpg"),(400,400),cv2.INTER_LINEAR)
ura_count *= 0.1
img2[:,:,:] = img2[:,:,:]*(1-ura_count) + img3[:,:,:] * ura_count
def check_(input_frame):
global HSV
hsv = cv2.cvtColor(input_frame,cv2.COLOR_BGR2HSV)
# Filtering pixels with Yellow Hue. (20~30)
idx = ((hsv[:,:,0]>20)*(hsv[:,:,0]<30)*(hsv[:,:,1]>90))
HSV = hsv.copy()
#here needs tuning of the threshold
if idx.sum() > 60000:
return True
# cv2.imshow("f",frame)
# function for detecting expression of Emoji
# Using AKAZE algorithm
def detect(face_name, img2):
global detected
global detected_counts
global faces
# reading template imgs
img1 = cv2.imread("img2/" + face_name + ".jpeg")
kp1, des1 = akaze.detectAndCompute(img1, None)
kp2, des2 = akaze.detectAndCompute(img2, None)
# BFM 生成
bf = cv2.BFMatcher()
# 特徴量ベクトル同士を BFM＆KNN でマッチング
if des2 is not None and des2.shape[0] > 1:
matches = bf.knnMatch(des1, des2, k=2)
# 全部だと激重になるのでデータを間引きする
# ratioが低いほど厳しい判定になる
ratio = 0.6
good = []
for data in matches:
if len(data) > 1:
m = data[0]
n = data[1]
if m.distance < ratio * n.distance:
good.append([m])
#もし特徴点が3つ以上あったら
if len(good) > 3:
print(face_name+" Existing!!!!!")
detected = face_name
detected_counts[faces.index(face_name)] += 1
return True
else: return False
# detect
def main():
global detected
global detected_counts
#デフォルトではnothing を表示しておく
detected = "nothing"
count = 0
ura_count = 0
time_before = time.time()
found_flg = False
#何秒ごとにdetect関数を呼ぶか
interval = 0.5
while(cap.isOpened()):
# print(detected_counts)
if detected == "nothing":
detected = "nothing"
ret, frame = cap.read()
found_flg = False
if frame is not None:
cv2.rectangle(frame, (int(cap_w/2 - window_width/2), int(cap_h/2 - window_height/2)), (int(cap_w/2 + window_width/2), int(cap_h/2 + window_height/2)), (0,0,255), thickness = 2)
img2 = frame[int(cap_h/2 - window_height/2):int(cap_h/2 + window_height/2),int(cap_w/2 - window_width/2):int(cap_w/2 + window_width/2)]
# 特徴量の検出と特徴量ベクトルの計算
time_after = time.time()
#interval秒経ったら
if time_after - time_before > interval:
time_before = time.time()
for face in faces:
if found_flg == False:
found_flg = detect(face, img2)
else:
break
if found_flg != True:
detected = "nothing"
if check_(img2):
#ura_count is for counting the number of continuous backface of Emoji detection
ura_count = min(ura_count + 2, 7)
print("Ura!?!?!?!?")
else:
ura_count = max(0, ura_count - 3)
else:
ura_count = max(0, ura_count - 3)
display_face(ura_count,img2)
if ura_count > 3:
cv2.putText(frame, "Ura!!!!!", (100, int(cap_h) - 100), cv2.FONT_HERSHEY_SIMPLEX, 2.0, (255, 255, 255), thickness=3)
cv2.putText(frame, detected, (int(cap_w) - 300, int(cap_h) - 100), cv2.FONT_HERSHEY_SIMPLEX, 2.0, (255, 255, 255), thickness=3)
cv2.namedWindow("frame", cv2.WINDOW_NORMAL)
cv2.imshow('frame', frame)
if cv2.waitKey(1) & 0xFF == ord('q'):
break
cap.release()
cv2.destroyAllWindows()
if __name__ == '__main__':
my_capture()
main()
# hsv = cv2.cvtColor(frame,cv2.COLOR_BGR2HSV)
# idx = np.logical_not((hsv[:,:,0]>20)*(hsv[:,:,0]<30)*(hsv[:,:,1]>90))
# frame[idx,:] = 255