【深度学习实战—番外篇2】：如何保存CIFAR-10数据集图片？????一、简述????二、程序编写????三、图片可视化

96 阅读 0 评论 64 点赞

我是靠谱客的博主开放薯片，这篇文章主要介绍【深度学习实战—番外篇2】：如何保存CIFAR-10数据集图片？????一、简述????二、程序编写????三、图片可视化，现在分享给大家，希望可以做个参考。

✨博客主页：米开朗琪罗~????
✨博主爱好：羽毛球????
✨年轻人要：Living for the moment（活在当下）！????
????推荐专栏：【图像处理】【千锤百炼Python】【深度学习】【排序算法】

????一、简述

CIFAR-10数据集是一个包含了60000张32×32的RGB图像数据集，所有图像共被划分为10个类别，分别为：

复制代码

1
2
3
['飞机', '汽车', '鸟', '猫', '鹿', '狗', '青蛙', '马', '船', '卡车']
['airplane', 'automobile', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']

【深度学习实战—2】：基于Keras的cifar10图像分类（非常详细、代码开源）这篇博文已经介绍了如何使用CIFAR-10数据集并基于keras框架做了一个分类任务！

该篇博文的目的就是如何将CIFAR-10数据集的图片导出并保存至本地文件夹中。????

????二、程序编写

????2.1 文件夹建立

步骤一：首先我们在桌面（也可为其他路径）路径下建立一个名为cifar10的空文件夹；
在这里插入图片描述
步骤二：在cifar10文件夹下建立train和test文件夹；

步骤三：分别在train和test文件夹下建立10个与图像类别名称相同的空文件夹。

????2.2 程序编写

对训练集数据进行图像保存的程序如下所示：

复制代码

from keras.datasets import cifar10
from PIL import Image

"""
数据集下载与加载（利用KerasAPI）
"""
# x_train_original和y_train_original代表训练集的图像与标签, x_test_original与y_test_original代表测试集的图像与标签
(x_train_original, y_train_original), (x_test_original, y_test_original) = cifar10.load_data()

root_path = 'C:\Users\Lenovo\Desktop\cifar10\'

def mnist_train_save():
    for i in range(60000):
        print(i)
        if (y_train_original[i] == 0):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\airplane\' + str(i) + '.jpg')
        elif (y_train_original[i] == 1):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\automobile\' + str(i) + '.jpg')
        elif (y_train_original[i] == 2):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\bird\' + str(i) + '.jpg')
        elif (y_train_original[i] == 3):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\cat\' + str(i) + '.jpg')
        elif (y_train_original[i] == 4):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\deer\' + str(i) + '.jpg')
        elif (y_train_original[i] == 5):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\dog\' + str(i) + '.jpg')
        elif (y_train_original[i] == 6):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\frog\' + str(i) + '.jpg')
        elif (y_train_original[i] == 7):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\horse\' + str(i) + '.jpg')
        elif (y_train_original[i] == 8):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\ship\' + str(i) + '.jpg')
        else:
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\truck\' + str(i) + '.jpg')

mnist_train_save()
print('图片保存完成')

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
from keras.datasets import cifar10
from PIL import Image

"""
数据集下载与加载（利用KerasAPI）
"""
# x_train_original和y_train_original代表训练集的图像与标签, x_test_original与y_test_original代表测试集的图像与标签
(x_train_original, y_train_original), (x_test_original, y_test_original) = cifar10.load_data()

root_path = 'C:\Users\Lenovo\Desktop\cifar10\'


def mnist_train_save():
    for i in range(60000):
        print(i)
        if (y_train_original[i] == 0):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\airplane\' + str(i) + '.jpg')
        elif (y_train_original[i] == 1):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\automobile\' + str(i) + '.jpg')
        elif (y_train_original[i] == 2):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\bird\' + str(i) + '.jpg')
        elif (y_train_original[i] == 3):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\cat\' + str(i) + '.jpg')
        elif (y_train_original[i] == 4):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\deer\' + str(i) + '.jpg')
        elif (y_train_original[i] == 5):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\dog\' + str(i) + '.jpg')
        elif (y_train_original[i] == 6):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\frog\' + str(i) + '.jpg')
        elif (y_train_original[i] == 7):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\horse\' + str(i) + '.jpg')
        elif (y_train_original[i] == 8):
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\ship\' + str(i) + '.jpg')
        else:
            img = Image.fromarray(x_train_original[i])
            img.save(root_path + 'train\truck\' + str(i) + '.jpg')


mnist_train_save()
print('图片保存完成')