-
Notifications
You must be signed in to change notification settings - Fork 3
/
Copy pathmain.py
82 lines (75 loc) · 3.71 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
import cv2
import numpy as np
import dlib
#人脸检测
detector = dlib.get_frontal_face_detector()
#人脸landmark
PREDICTOR_PATH = "shape_predictor_68_face_landmarks.dat"
predictor = dlib.shape_predictor(PREDICTOR_PATH)
#猫脸检测
cat_path = ".\haarcascade_frontalcatface.xml"
facecascade = cv2.CascadeClassifier(cat_path)
ret = facecascade.load(cat_path)
#猫脸landmark，这个没有搜到源码，暂时不做
old_cat_face_loc = np.array([-1,-1])
def human_mouth_paste_to_cat(human,cat):
global old_cat_face_loc
human_gray = cv2.cvtColor(human,cv2.COLOR_BGR2GRAY)
points_keys = []
#人脸检测
rects = detector(human_gray,1)
#人脸landmark检测
landmarks = np.matrix([[p.x,p.y] for p in predictor(human_gray,rects[0]).parts()])
landmarks = np.array(landmarks)
#mouth的landmark
mouth_landmark = landmarks[48:,:]
#扩个边
border=8
mouth = human[np.min(mouth_landmark[:,1])-border:np.max(mouth_landmark[:,1]+border),np.min(mouth_landmark[:,0])-border:np.max(mouth_landmark[:,0])+border,:]
mouth_landmark[:,0] -= (np.min(mouth_landmark[:,0])-border)
mouth_landmark[:,1] -= (np.min(mouth_landmark[:,1])-border)
#制作用于泊松融合的mask
mask=np.zeros((mouth.shape[0],mouth.shape[1],3)).astype(np.float32)
for i in range(mouth_landmark.shape[0]):#先画线
cv2.line(mask,(mouth_landmark[i,0],mouth_landmark[i,1]),(mouth_landmark[(i+1)%mouth_landmark.shape[0],0],mouth_landmark[(i+1)%mouth_landmark.shape[0],1]),(255,255,255),10)
mask_tmp=mask.copy()
for i in range(6,mask.shape[0]-6):#将线内部的范围都算作mask=255
for j in range(6,mask.shape[1]-6):
if(np.max(mask_tmp[:i,:j,:])==0 or np.max(mask_tmp[i:,:j,:])==0 or np.max(mask_tmp[:i,j:,:])==0 or np.max(mask_tmp[i:,j:,:])==0):
mask[i,j,:]=0
else:
mask[i,j,:]=255
#猫脸检测
width, height, channels = cat.shape
cat_gray = cv2.cvtColor(cat,cv2.COLOR_BGR2GRAY)
cat_face_loc= facecascade.detectMultiScale(cat_gray,scaleFactor = 1.1,minNeighbors=3,minSize=(100,100),flags=cv2.CASCADE_SCALE_IMAGE)
cat_face_loc = cat_face_loc[0]
if(old_cat_face_loc[0] != -1):#因为猫脸检测抖动太厉害，所以此处用历史坐标缓冲一下
cat_face_loc = 0.9*old_cat_face_loc + 0.1*cat_face_loc
old_cat_face_loc = cat_face_loc
center = (int(cat_face_loc[0]+cat_face_loc[2]/2), int(cat_face_loc[1]+cat_face_loc[3]*0.8))#0.8为手动设定的猫嘴位置，因为没找到猫脸landmark
normal_clone = cv2.seamlessClone(mouth, cat, mask.astype(mouth.dtype), center, cv2.NORMAL_CLONE)
#mixed_clone = cv2.seamlessClone(mouth, cat, mask.astype(mouth.dtype), center, cv2.MIXED_CLONE)
#cv2.imwrite("opencv-normal-clone-example.jpg", normal_clone)
#cv2.imwrite("opencv-mixed-clone-example.jpg", mixed_clone)
return normal_clone
#human = cv2.imread('bigmouth.jpg')
#cat = cv2.imread('cat.jpg')
#cat_with_human_mouth = human_mouth_paste_to_cat(human,cat)
human_video_cap = cv2.VideoCapture("human_video.mp4")
cat_video_cap =cv2.VideoCapture("cat_video2.mp4")
fourcc = cv2.VideoWriter_fourcc(*'MJPG')
video_writer = cv2.VideoWriter('cat_with_humanmouth.avi', fourcc, 29.43, (720,1280))
index=0
while(True):
print(index)
index+=1
if(index>400):
break
human_ret,human_frame = human_video_cap.read()
human_frame = cv2.resize(human_frame,dsize=None,fx=2,fy=2)
cat_ret,cat_frame = cat_video_cap.read()
if(human_ret == True and cat_ret == True):
cat_with_human_mouth = human_mouth_paste_to_cat(human_frame,cat_frame)
video_writer.write(cat_with_human_mouth.astype(np.uint8))
video_writer.release()