Algae_detect/dataloader.py at master · aoxipo/Algae_detect · 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
145
146
147
148
149
150
151
152
153
154
155
156
import random
import numpy as np
import os
import torch
from torch.utils.data import Dataset, DataLoader, TensorDataset
from torch.autograd import Variable
#from sklearn.model_selection import train_test_split
import torchvision
import torchvision.transforms as transforms
from PIL import Image
import cv2
import matplotlib.pyplot as plt
import glob


class Dataload(Dataset):
    def __init__(self, file_path, batch_size = 1, data_source = None, gray = False, image_shape = (224,224)
                 , same_matrix = True, num_require = 25,data_type = 'train'):
        self.file_path = file_path
        self.data_source = data_source
        self.batch_size = batch_size
        self.image_shape = image_shape
        self.data_type = data_type
        self.label_dict = {
            0:'Bacillariophyta',
            1:'Chlorella',
            2:'Chrysophyta',
            3:'Dunaliella_salina',
            4:'Platymonas',
            5:'translating_Symbiodinium',
            6:'bleaching_Symbiodinium',
            7:'normal_Symbiodinium}'
        }
        self.num_class = len(self.label_dict)
        self.photo_set = []
        self.same_matrix =same_matrix
        self.gray = gray
        self.num_require = num_require
        self.load_data(file_path)
        self.set_gan()

        # self.X_train, self.Y_train = self.load_all_data(False ,gray, "train")
        # self.X_val, self.Y_val = self.load_all_data(False ,gray, "val")

    def check_dir(self, path):
        if (not os.path.exists(path)):
            return 0
        return 1

    def read_image_data(self, file_path, gray = False):
        if(gray):
            image = cv2.imread(file_path, 0)
        else:
            image = cv2.imread(file_path)
        if(image is None):
             raise RuntimeError('image can \'t read:' + file_path)
        return image

    def set_gan(self):
        cifar_norm_mean = (0.49139968, 0.48215827, 0.44653124)
        cifar_norm_std = (0.24703233, 0.24348505, 0.26158768)
        method_list = [
                transforms.ToPILImage(),
                transforms.Resize(self.image_shape),
                transforms.ToTensor(),
                transforms.Normalize(cifar_norm_mean, cifar_norm_std),
        ]
        self.datagen = transforms.Compose(method_list)

    def load_data(self, file_path):

        middle = file_path + "\\images\\"
        label_path = file_path + "\\labels\\"

        for i in os.listdir(middle):
            num = os.path.splitext(i)[0]
            x = middle + num + '.png'
            y = label_path + num + '.txt'
            self.photo_set.append([x,y])

        if(not self.check_dir(middle)):
            if(self.dataset_type == "train"):
                raise RuntimeError('train dir not exists:'+ middle)
            else:
                raise RuntimeError('val dir not exists:' + middle)
        self.total_number = len(self.photo_set)
        print("total:", self.total_number)

    def __getitem__(self, index):
        """
        获取对应index的图像，并视情况进行数据增强
        """
        if(index >= self.total_number):
            raise StopIteration
        try:
            re_index = index
            if(len(self.photo_set)>0):
                image_path = self.photo_set[re_index][0]
                image = self.read_image_data(image_path,self.gray)
                label = []
                if(self.data_type == "train"):
                    with open(self.photo_set[re_index][1]) as f:
                        lines = f.readlines()
                        for line in lines:
                            x = line.replace('\n', '').split(' ')
                            x = [float(i) for i in x]
                            label.append(x)
                    l = len(label)
                    if(self.same_matrix):
                        ones = np.zeros([self.num_require, 5])
                        ones[:,0] = self.num_class
                        ones[:l,:] =label
                        label = ones
                else:
                    label = np.zeros([self.num_require, 5])

            label = torch.tensor(label)
            if self.datagen is not None:
                image = self.datagen(image)

            return image, label

        except Exception as e:
            print("发现异常")
            print(e.__class__.__name__)
            print(e)


    def __len__(self):
        return len(self.photo_set)


if __name__ == '__main__':

    batch_size = 32

    train_dataloader = Dataload(r"E:\Dataset\training_set\train")

    train_loader= DataLoader(
        dataset = train_dataloader,
        batch_size = batch_size,
        shuffle = False,
        drop_last = True
    )
    # re_index, b= train_loader[500]
    # print("photo")
    # print(re_index)
    # print("label")
    # print(b)

    for data in train_loader:
        img,label =data
        print(img.shape)