forked from tensorflower/seetaFace6Python
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathface_landmark_demo.py
103 lines (91 loc) · 3.95 KB
/
face_landmark_demo.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
#_*_coding:utf-8 _*_
"""
人脸关键点检测模型
当前主要有：
1. 5点人脸检测
2. 68点人脸检测
3. 带有关键点是否遮挡判断的的5点人脸检测
"""
from seetaface.api import *
"""
总共有三种关键点检测模型
1.普通5点检测模型
要加载的功能 :
LANDMARKER5：普通5点检测模型
依赖功能:
FACE_DETECT：人脸检测
使用到的函数:
mark5：检测人脸5个关键点位置
2.普通68点检测模型
要加载的功能 :
LANDMARKER68：普通68点检测模型
依赖功能:
FACE_DETECT：人脸检测
使用到的函数:
mark68：检测人脸68个关键点位置
3.带关键点检测被遮挡判断的5点检测模型
要加载的功能 :
LANDMARKER_MASK：带关键点检测遮挡判断的5点检测模型
依赖功能:
FACE_DETECT：人脸检测
使用到的函数:
markMask：检测人脸5个关键点位置，和这5点是否被遮挡的判断
"""
init_mask = FACE_DETECT|LANDMARKER5|LANDMARKER68|LANDMARKER_MASK
seetaFace = SeetaFace(init_mask)
#########################################################
################ 5 点检测模型 ################
#########################################################
points_5_image = cv2.imread("asserts/1.jpg")
detect_result = seetaFace.Detect(points_5_image)
for i in range(detect_result.size):
face = detect_result.data[i].pos
#5点检测模型检测
points_5 = seetaFace.mark5(points_5_image,face)
#画人脸框
cv2.rectangle(points_5_image, (face.x, face.y), (face.x + face.width, face.y + face.height), (255, 0, 0),2)
for i in range(5):
print(points_5[i])
# 画关键点
cv2.circle(points_5_image, (int(points_5[i].x), int(points_5[i].y)), 5, (255, 0, 0), -1)
#########################################################
################ 68点检测模型 ################
#########################################################
points_68_image = cv2.imread("asserts/1.jpg")
detect_result = seetaFace.Detect(points_68_image)
for i in range(detect_result.size):
face = detect_result.data[i].pos
# 68点检测模型检测
points_68 = seetaFace.mark68(points_68_image, face)
#画人脸框
cv2.rectangle(points_68_image, (face.x, face.y), (face.x + face.width, face.y + face.height), (255, 0, 0),2)
for i in range(68):
print(points_68[i])
# 画关键点
cv2.circle(points_68_image, (int(points_68[i].x), int(points_68[i].y)), 3, (255, 0, 0), -1)
#########################################################
#######带关键点是否遮挡的 5点检测模型检测 ###########
#########################################################
points_mask_image = cv2.imread("asserts/mask.png")
detect_result = seetaFace.Detect(points_mask_image)
for i in range(detect_result.size):
face = detect_result.data[i].pos
# 带关键点是否遮挡的 5点检测模型检测，返回两个值 （points,face_mask），
# point 基本等价于5点检测模型中的关键点位置数据，face_mask 为一个 长度为5的数组，里面记录每个关键点是否被遮挡
#如 戴口罩时 face_mask 可能会返回 [左眼，右眼，鼻子，左边嘴角，右边嘴角] ->[0,0,1,1,1] ,0：没被遮挡 1：被遮挡
points_mask = seetaFace.markMask(points_mask_image, face)
#画人脸框
cv2.rectangle(points_mask_image, (face.x, face.y), (face.x + face.width, face.y + face.height), (255, 0, 0),2)
points,mask = points_mask
for i in range(5):
print(points[i])
if mask[i]==1:#被遮挡
#画关键点
cv2.circle(points_mask_image, (int(points[i].x), int(points[i].y)), 5, (0, 0, 255), -1)
else:
# 画关键点
cv2.circle(points_mask_image, (int(points[i].x), int(points[i].y)), 5, (255, 0, 0), -1)
cv2.imshow("points_5_image",points_5_image)
cv2.imshow("points_68_image",points_68_image)
cv2.imshow("points_mask_image",points_mask_image)
cv2.waitKey(0)