-
Notifications
You must be signed in to change notification settings - Fork 1
/
data_process.py
112 lines (96 loc) · 6.52 KB
/
data_process.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
import os
from utils.img_read_save import image_read_cv2
import h5py
import numpy as np
from tqdm import tqdm
import cv2
img_text_path = 'VLFDataset'
h5_path = "VLFDataset_h5"
os.makedirs(h5_path, exist_ok=True)
task_name = 'IVF'
dataset_name = 'MSRS'
dataset_mode = 'train'
size = 'small'
h5_file_path = os.path.join(h5_path, dataset_name + '_' + dataset_mode +'.h5')
small_size_0 = (384, 288)
small_size_1 = (288, 384)
with h5py.File(h5_file_path, 'w') as h5_file:
# 创建两个数据集组
imageA = h5_file.create_group('imageA')
imageB = h5_file.create_group('imageB')
text = h5_file.create_group('text')
text_txt = os.path.join(img_text_path, 'Image', task_name, dataset_name, dataset_mode + '.txt')
with open(text_txt, 'r') as file:
file_list = [line.strip() for line in file.readlines()]
sample_names = []
for name in file_list:
name = name.split('.')[0]
sample_names.append(name)
# 获取所有样本的文件名（不包括文件扩展名）
if task_name == 'MFF':
# 遍历每个样本
for sample_name in tqdm(sample_names):
if dataset_mode == 'train' and size == 'small':
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'MRI', sample_name + '.png'), mode='GRAY')
h, w = img_A.shape
if h < w:
img_A = cv2.resize(img_A, small_size_0)[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'CT', sample_name + '.png'), mode='GRAY'), small_size_0)[None, ...] / 255.0
else:
img_A = cv2.resize(img_A, small_size_1).T[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'CT', sample_name + '.png'), mode='GRAY'), small_size_1).T[None, ...] / 255.0
else:
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'NEAR', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
img_B = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'FAR', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
text_feature = np.load(os.path.join(img_text_path, 'TextFeature', task_name, dataset_name, sample_name + '.npy'))
# 将图像和文本保存到HDF5文件中
imageA.create_dataset(sample_name, data=img_A)
imageB.create_dataset(sample_name, data=img_B)
text.create_dataset(sample_name, data=text_feature)
elif task_name == 'MEF':
for sample_name in tqdm(sample_names):
if dataset_mode == 'train' and size == 'small':
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'OVER', sample_name + '.png'), mode='GRAY')
h, w = img_A.shape
if h < w:
img_A = cv2.resize(img_A, small_size_0)[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'UNDER', sample_name + '.png'), mode='GRAY'), small_size_0)[None, ...] / 255.0
else:
img_A = cv2.resize(img_A, small_size_1).T[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'UNDER', sample_name + '.png'), mode='GRAY'), small_size_1).T[None, ...] / 255.0
else:
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'OVER', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
img_B = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'UNDER', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
text_feature = np.load(os.path.join(img_text_path, 'TextFeature', task_name, dataset_name, sample_name + '.npy'))
# 将图像和文本保存到HDF5文件中
imageA.create_dataset(sample_name, data=img_A)
imageB.create_dataset(sample_name, data=img_B)
text.create_dataset(sample_name, data=text_feature)
elif task_name == 'IVF':
for sample_name in tqdm(sample_names):
if dataset_mode == 'train' and size == 'small':
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'IR', sample_name + '.png'), mode='GRAY')
h, w = img_A.shape
if h < w:
img_A = cv2.resize(img_A, small_size_0)[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'VI', sample_name + '.png'), mode='GRAY'), small_size_0)[None, ...] / 255.0
else:
img_A = cv2.resize(img_A, small_size_1).T[None, ...] / 255.0
img_B = cv2.resize(image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'VI', sample_name + '.png'), mode='GRAY'), small_size_1).T[None, ...] / 255.0
else:
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'IR', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
img_B = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'VI', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
text_feature = np.load(os.path.join(img_text_path, 'TextFeature', task_name, dataset_name, sample_name + '.npy'))
# 将图像和文本保存到HDF5文件中
imageA.create_dataset(sample_name, data=img_A)
imageB.create_dataset(sample_name, data=img_B)
text.create_dataset(sample_name, data=text_feature)
elif task_name == 'MIF':
for sample_name in tqdm(sample_names):
img_A = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'MRI', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
img_B = image_read_cv2(os.path.join(img_text_path, 'Image', task_name, dataset_name, 'T', sample_name + '.png'), mode='GRAY')[None, ...] / 255.0
text_feature = np.load(os.path.join(img_text_path, 'TextFeature', task_name, dataset_name, sample_name + '.npy'))
# 将图像和文本保存到HDF5文件中
imageA.create_dataset(sample_name, data=img_A)
imageB.create_dataset(sample_name, data=img_B)
text.create_dataset(sample_name, data=text_feature)