-
Notifications
You must be signed in to change notification settings - Fork 1
/
data_image_helper.py
202 lines (169 loc) · 7.13 KB
/
data_image_helper.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
# -*- coding: utf-8 -*-
"""
Created on Mon Jan 14 20:55:34 2019
@author: 50568
"""
import data_file_helper as fh
import tensorflow as tf
# import tensorflow.contrib as tf_contrib
import numpy as np
import cv2
import visualization
class data_image_helper:
def __init__(self, detector):
self.detector = detector
def read_img(self, path, shape, size, begin=0, end=0):
"""
Video_Read is used to extract the image of mouth from a video;\n
parameter:\n
Path: the string path of video\n
Shape: the (min, max) size tuple of the mouth you extract from the video\n
Size: the (high, weight) size tuple of the mouth image you save
"""
cap = cv2.VideoCapture(path)
images = []
mouth = None
cnt = 0
frames = cap.get(cv2.CAP_PROP_FRAME_COUNT)
fps = cap.get(cv2.CAP_PROP_FPS)
v_length = frames / fps
if (end == 0 or end >= v_length):
end = v_length
if (cap.isOpened() == False):
print("Read video failed!")
return None
# get detector
classifier_face = cv2.CascadeClassifier(
"./cascades/haarcascade_frontalface_alt.xml")
classifier_mouth = cv2.CascadeClassifier(
"./cascades/haarcascade_mcs_mouth.xml")
cap.set(cv2.CAP_PROP_POS_FRAMES, begin * fps)
pos = cap.get(cv2.CAP_PROP_POS_FRAMES)
while (pos <= end * fps and end <= v_length):
ret, img = cap.read()
'''
第一个参数ret的值为True或False，代表有没有读到图片
第二个参数是frame，是当前截取一帧的图片
'''
pos = cap.get(cv2.CAP_PROP_POS_FRAMES)
if ret == False:
break
faceRects_face = classifier_face.detectMultiScale(
img, 1.2, 2, cv2.CASCADE_SCALE_IMAGE, (20, 25))
key = cv2.waitKey(1)
# 键盘等待
if len(faceRects_face) > 0:
# 检测到人脸
for faceRect_face in faceRects_face:
# 获取图像x起点,y起点,宽，高
x, y, w, h = faceRect_face
# 转换类型为int，方便之后图像截取
intx = int(x)
intw = int(w)
# 截取人脸区域下半部分左上角的y起点，以精确识别嘴巴的位置
my = int(float(y + 0.6 * h))
mh = int(0.5 * h)
img_facehalf_bottom = img[my:(my + mh), intx:intx + intw]
'''
img获取坐标为，【y,y+h之间（竖）：x,x+w之间(横)范围内的数组】
img_facehalf是截取人脸识别到区域上半部分
img_facehalf_bottom是截取人脸识别到区域下半部分
'''
cv2.rectangle(img, (int(x), my),
(int(x) + int(w), my + mh), (0, 255, 0), 2,
0)
'''
矩形画出区域 rectangle参数（图像，左顶点坐标(x,y)，右下顶点坐标（x+w,y+h），线条颜色，线条粗细）
画出人脸识别下部分区域，方便定位
'''
faceRects_mouth = classifier_mouth.detectMultiScale(
img_facehalf_bottom, 1.1, 1, cv2.CASCADE_SCALE_IMAGE,
shape)
if len(faceRects_mouth) > 0:
for faceRect_mouth in faceRects_mouth:
xm1, ym1, wm1, hm2 = faceRect_mouth
cv2.rectangle(
img_facehalf_bottom, (int(xm1), int(ym1)),
(int(xm1) + int(wm1), int(ym1) + int(hm2)),
(0, 0, 255), 2, 0)
mouth = img_facehalf_bottom[ym1:(ym1 + hm2), xm1:(
xm1 + wm1)]
mouth = cv2.resize(
mouth, size, interpolation=cv2.INTER_CUBIC)
images.append(mouth)
cnt += 1
# cv2.imshow('video', mouth)
# if cnt % 10 == 0:
# cv2.imwrite(str(cnt) + 'xx.jpg', mouth)
if (key == ord('q')):
break
cap.release()
cv2.destroyAllWindows()
return images, cnt
# def prepare_data(self,
# path,
# batch_size,
# time_step,
# shape = (20, 20),
# size = (109, 109),
# read = True):
# if(read):
# self.read_img(path, shape, size)
# DataSet = []
# Buffers = [None] * time_step
# cnt = 0
# for image in self.images:
# cnt += 1
# for i in range(time_step):
# Buffers[time_step -i - 1] = Buffers[time_step - i - 2]
# Buffers[0] = image
# if(cnt >= time_step):
# DataSet.append(Buffers.copy())
# # DataSet = DataSet / 255.0
# # DataSet = DataSet.astype(np.float32)
# batch_dataset = tf.data.Dataset.from_tensor_slices(DataSet)
# batch_dataset = batch_dataset.batch(batch_size)
# return batch_dataset, self.images
def get_raw_dataset(self, path, shape=(20, 20), size=(224, 224)):
video, cnt = self.read_img(path, shape, size, 0.5, 1)
video = np.array(video) / 255.0
video = video.astype(np.float32)
return video
# return tf.data.Dataset.from_generator(generator, tf.float32)
def prepare_data(
self,
paths,
batch_size,
shape=(20, 20),
size=(224, 224),
):
dataset = []
length = []
for path in paths:
video, cnt = self.read_img(path, shape, size, 0.5, 1)
video = np.array(video) / 255.0
video = video.astype(np.float32)
dataset.append(video)
length.append(cnt)
def generator():
for d, c in zip(dataset, length):
yield d, c
raw_dataset = tf.data.Dataset.from_generator(generator,
(tf.float32, tf.int32))
batch_dataset = raw_dataset.padded_batch(
batch_size,
padded_shapes=(tf.TensorShape([None, 109, 109, 5]),
tf.TensorShape([])))
return batch_dataset, raw_dataset
if __name__ == '__main__':
video, txt = fh.read_file(
'/Users/barid/Documents/workspace/batch_data/lip_data')
print(video[:5])
print(txt[:5])
helper = data_image_helper(detector='./cascades/')
# b, d = helper.prepare_data(paths = ['D:/lip_data/ABOUT/train/ABOUT_00003.mp4'], batch_size = 64)
b, d = helper.prepare_data(paths=video, batch_size=32)
print(b)
#for (i,(x, l)) in enumerate(b):
# print(x)
# print(l)