深度学习数据集的文本制作和读取

制作数据集的文本文件

import os
from os.path import join
import random
import config
args = config.args

class SplitDataset:
    def __init__(self):
        self.data_root_path = args.data_root_path
        self.dataset_split_rate = args.dataset_split_rate

    def split_dataset(self):
        # image_file_list里面只保留有标签数据的
        label_path = join(self.data_root_path, 'label')
        image_path = join(self.data_root_path, 'image')
        dataset_name_list = []
        for image_name in os.listdir(image_path):
            label_name = image_name.split(sep='.')[0]+"_GT.nii"
            if label_name in os.listdir(label_path):
                dataset_name_list.append(join(image_name))
        # print(image_name_list)
        # 将image_file_list分为训练、验证和测试集
        numbers = len(dataset_name_list)
        print('Total numbers of samples is :', numbers)
        random.shuffle(dataset_name_list)
        train_name_list = dataset_name_list[0:int(numbers * self.dataset_split_rate[0])]
        val_name_list = dataset_name_list[
                        int(numbers * self.dataset_split_rate[0]):
                        int(numbers * (self.dataset_split_rate[0] + self.dataset_split_rate[1]))]
        test_name_list = dataset_name_list[
                        int(numbers * (self.dataset_split_rate[0] + self.dataset_split_rate[1])):]

        self.write_txt(train_name_list, "train_list.txt")
        self.write_txt(val_name_list, "val_list.txt")
        self.write_txt(test_name_list, "test_list.txt")

    def write_txt(self, name_list, file_name):
        root_path = join(args.root_path, 'TOF_MRA')
        f = open(join(root_path, file_name), 'w')
        for name in name_list:
            image_path = join(args.data_root_path, 'image', name)
            label_path = join(args.data_root_path, 'label', name.split(sep='.')[0]+'_GT.nii')
            f.write(image_path + ' ' + label_path + "\n")
        f.close()


if __name__ == '__main__':
    dataset = SplitDataset().split_dataset()
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

结果

读取文本文件

def load_file_name_list(file_path):
    file_name_list = []
    with open(file_path, 'r') as file_to_read:
        while True:
            lines = file_to_read.readline().strip()  # 整行读取数据
            if not lines:
                break
                pass
            file_name_list.append(lines)
            pass
    return file_name_list
1
2
3
4
5
6
7
8
9
10
11