pytorch创建数据集

最新推荐文章于 2024-09-05 11:42:07 发布

oldsweet老甜甜

最新推荐文章于 2024-09-05 11:42:07 发布

阅读量116

点赞数

文章标签： python 深度学习人工智能

本文链接：https://blog.csdn.net/qq_45867029/article/details/117752444

版权

from torch.utils.data import Dataset
import cv2
import os


class MyData(Dataset):
    def __init__(self, root_dir, label_dir):

        self.root_dir = root_dir
        self.label_dir = label_dir
        self.path = os.path.join(root_dir, label_dir)
        self.image_dir_list = os.listdir(self.path)

    def __getitem__(self, item):
        image_dir = os.path.join(self.path, item)
        image = cv2.imread(image_dir)
        label = self.label_dir
        return image, label

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


if __name__ == '__main__':
    mydata = MyData(root_dir="hymenoptera_data/train", label_dir="ants")
    image_dir_list = mydata.image_dir_list
    for index in image_dir_list:
        cv2.imshow(mydata.__getitem__(index)[1],mydata.__getitem__(index)[0])
        cv2.waitKey(1000)
        cv2.destroyAllWindows()