deepupe_tensorflow/dataset.py at master · SystemErrorWang/deepupe_tensorflow · GitHub

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
import cv2
import os
import numpy as np
from tqdm import tqdm
from torch.utils.data import Dataset, DataLoader
from torchvision import transforms


class HDRDataset(Dataset):


    def __init__(self, data_dir, mode='train', transform=None):
        total_len = len(os.listdir(data_dir))//2
        self.data_dir = data_dir
        if mode == 'train':
            self.data_len = 3000

        elif mode == 'test':
            self.data_len = total_len - 3000

        self.transform = transform


    def __len__(self):
        return self.data_len


    def __getitem__(self, idx):
        image_path = os.path.join(self.data_dir, '{}.jpg'.format(str(idx).zfill(4)))
        label_path = os.path.join(self.data_dir, '{}_gt.jpg'.format(str(idx).zfill(4)))
        image = cv2.imread(image_path)
        label = cv2.imread(label_path)
        h1, w1 = np.shape(image)[:2]
        h2, w2 = np.shape(label)[:2]

        if h1 > h2:
            dh = (h1-h2)//2
            image = image[dh: dh+h2, :, :]
        elif h1 < h2:
            dh = (h2-h1)//2
            label = label[dh: dh+h1, :, :]

        if w1 > w2:
            dw = (w1-w2)//2
            image = image[:, dw: dw+w2, :]
        elif w1 < w2:
            dw = (w1-w2)//2
            label = label[:, dw: dw+w1, :]

        try:
            assert np.shape(image) == np.shape(label)
        except:
            print(np.shape(image), np.shape(label))


        if self.transform is not None:
            image, label = self.transform(image, label)

        return image, label


class TrainTransform():

    def __init__(self, output_size):
        assert isinstance(output_size, (int, tuple))
        if isinstance(output_size, int):
            self.output_size = (output_size, output_size)
        else:
            assert len(output_size) == 2
            self.output_size = output_size


    def __call__(self, image, label):

        new_h, new_w = self.output_size

        h, w = np.shape(image)[:2]
        offset_h = np.random.randint(0, h - new_h)
        offset_w = np.random.randint(0, w - new_w)

        image = image[offset_h: offset_h + new_h,
                                offset_w: offset_w + new_w]
        label = label[offset_h: offset_h + new_h,
                                offset_w: offset_w + new_w]

        flip_prop = np.random.randint(0, 100)
        if flip_prop > 50:
            image = cv2.flip(image, 1)
            label = cv2.flip(label, 1)

        image = image.astype(np.float32)/255.0
        label = label.astype(np.float32)/255.0


        return image, label


class TestTransform():

    def __init__(self):
        pass


    def __call__(self, image, label):

        image = image.astype(np.float32)/255.0
        #label = label.astype(np.float32)/127.5 - 1

        return image, label


def get_train_loader(image_size, batch_size, data_dir):
    transform = TrainTransform(image_size)
    dataset = HDRDataset(data_dir, transform=transform)
    dataloader = DataLoader(dataset, batch_size=batch_size,
                            shuffle=True, num_workers=8)
    return dataloader


def get_test_loader(data_dir):
    transform = TestTransform()
    dataset = HDRDataset(data_dir, mode='test', transform=transform)
    dataloader = DataLoader(dataset, batch_size=1,
                            shuffle=False, num_workers=0)
    return dataloader


if __name__ == '__main__':
    data_dir = '/media/wangxinrui/新加卷/hdr+burst/hdr_burst'
    dataloader = get_train_loader(512, 8, data_dir)
    for idx, batch in tqdm(enumerate(dataloader), total=len(dataloader)):
        print(np.shape(batch[0]), np.shape(batch[1]))
        #pass