PyTorch加载 数据集 PyTorch加载自己的数据集实例详解

from torch.utils.data import DataLoader,Dataset
from skimage import io,transform
import matplotlib.pyplot as plt
import oimport torch
from torchvision import transforms, utils
from PIL import Image
import pandas as pd
import numpy as np
#过滤警告信息
import warnings
warnings.filterwarnings("ignore")

class MyDataset(Dataset): #继承Dataset
 def __init__(self, path_dir, transform=None): #初始化一些属性
  self.path_dir = path_dir #文件路径,如'.\data\cat-dog'
  self.transform = transform #对图形进行处理，如标准化、截取、转换等
  self.images = os.listdir(self.path_dir)#把路径下的所有文件放在一个列表中
 
 def __len__(self):#返回整个数据集的大小
  return len(self.images)
 
 def __getitem__(self,index):#根据索引index返回图像及标签
  image_index = self.images[index]#根据索引获取图像文件名称
  img_path = os.path.join(self.path_dir, image_index)#获取图像的路径或目录
  img = Image.open(img_path).convert('RGB')# 读取图像
    
  # 根据目录名称获取图像标签（cat或dog）
  label = img_path.split('\\')[-1].split('.')[0]
  #把字符转换为数字cat-0，dog-1
  label = 1 if 'dog' in label else 0
  
  if self.transform is not None:
   img = self.transform(img)
  return img,label

dataset = MyDataset('.\data\cat-dog',transform=None)
img, label = dataset[0] #将启动魔法方法__getitem__(0)
print(type(img))
<class 'PIL.Image.Image'>

from torchvision import transforms as T
transform = T.Compose([
 T.Resize(224), # 缩放图片(Image)，保持长宽比不变，最短边为224像素
 T.CenterCrop(224), # 从图片中间切出224*224的图片
 T.ToTensor(), # 将图片(Image)转成Tensor，归一化至[0, 1]
 T.Normalize(mean=[.5, .5, .5], std=[.5, .5, .5]) # 标准化至[-1, 1]，规定均值和标准差
])

dataset = MyDataset('.\data\cat-dog',transform=transform)
for img, label in dataset: 
 print("图像img的形状{},标签label的值{}".format(img.shape, label))
 print("图像数据预处理后：\n",img)
 break

#使用DataLoader加载数据
dataloader = DataLoader(dataset,batch_size=4,shuffle=True)
for batch_datas, batch_labels in dataloader:
 print(batch_datas.size(),batch_labels.size())
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([4, 3, 224, 224]) torch.Size([4])
torch.Size([2, 3, 224, 224]) torch.Size([2])

import torchvision
import matplotlib.pyplot as plt
import numpy as np
%matplotlib inline
# 显示图像
def imshow(img):
 img = img / 2 + 0.5  # unnormalize
 npimg = img.numpy()
 plt.imshow(np.transpose(npimg, (1, 2, 0)))
 plt.show()
# 随机获取部分训练数据
dataiter = iter(dataloader)
images, labels = dataiter.next()
# 显示图像
imshow(torchvision.utils.make_grid(images))
# 打印标签
print(' '.join('%s' % ["小狗" if labels[j].item()==1 else "小猫" for j in range(4)]))

import torch
from torchvision import transforms, datasets
data_transform = transforms.Compose([
  transforms.RandomSizedCrop(224),
  transforms.RandomHorizontalFlip(),
  transforms.ToTensor(),
  transforms.Normalize(mean=[0.485, 0.456, 0.406],
        std=[0.229, 0.224, 0.225])
 ])
hymenoptera_dataset = datasets.ImageFolder(root='.\catdog\train',
           transform=data_transform)
dataset_loader = torch.utils.data.DataLoader(hymenoptera_dataset,