Tensorboard的使用——02单张图片add_image

1、实例化SummaryWriter，并指明日志存放路径

writer = SummaryWriter("logs")

2、打开图片

img_path = "data/train/ants_image/20935278_9190345f6b.jpg"
img = Image.open(img_path)

3、将图片转换成tensor类型

tensor_train = transforms.ToTensor()
tensor_img = tensor_train(img)

也可以使用transforms.Compose将多个transforms变化组合在一起
4、调用实例，add图片
add_image接收numpy array, torch tensor, or caffe2 blob这几种类型，所以导入图片前因转换图片类型

writer.add_image("tensor_img",tensor_img)

5、关闭writer

writer.close()

6、在terminal中使用

tensorboard --logdir=logs

代码如下：

from torchvision import transforms
from torch.utils.tensorboard import SummaryWriter
from PIL import Image

writer = SummaryWriter("logs2")

img_path = "data/train/ants_image/20935278_9190345f6b.jpg"
img = Image.open(img_path)

tensor_train = transforms.ToTensor()
tensor_img = tensor_train(img)

writer.add_image("tensor_img",tensor_img)
writer.close()

在terminal中使用：

tensorboard --logdir=logs

tensorboard ：