Pytorch学习03_TensorBoard使用02

Opencv读取图片，获得numpy型数据类型

复制图片的相对路径

目前这种type不适用，考虑用numpy类型

安装opencv，在pytorch环境下

pip install opencv-python

导入numpy

import numpy as np

将PIL类型的img转换为 NumPy 数组

img_array=np.array(img)

HWC三通道

H：高度 W：宽度 C：通道

复制代码

from torch.utils.tensorboard import SummaryWriter
import numpy as np
from PIL import Image

writer = SummaryWriter("logs")
image_path="dataset/train/ants_image/0013035.jpg"
img_PIL=Image.open(image_path)
img_array=np.array(img_PIL)
print(type(img_array))
print(img_array.shape)

writer.add_image("test",img_array,1,dataformats='HWC')


# for i in range(100):
#     writer.add_scalar("y=2x",3*i,i)

writer.close()

从PIL到numpy，需要在add_image()中指定shape中每一个数字/维表示的含义

终端运行

tensorboard --logdir=logs --port=6007

点击蓝色链接

点击"IMAGES"

来到

修改一下

使用另一张图片的路径，运行

复制代码

from torch.utils.tensorboard import SummaryWriter
import numpy as np
from PIL import Image

writer = SummaryWriter("logs")
image_path="dataset/train/ants_image/0013035.jpg"
img_PIL=Image.open(image_path)
img_array=np.array(img_PIL)
print(type(img_array))
print(img_array.shape)

# writer.add_image("test",img_array,1,dataformats='HWC')
writer.add_image("test",img_array,2,dataformats='HWC')


# for i in range(100):
#     writer.add_scalar("y=2x",3*i,i)

writer.close()

回到网站，进行刷新

刷新后

拖动滑轮进行图片查看

拖到左边后，可以看到之前的图片

更换标签

复制代码

from torch.utils.tensorboard import SummaryWriter
import numpy as np
from PIL import Image

writer = SummaryWriter("logs")
# image_path="dataset/train/ants_image/0013035.jpg"
image_path="dataset/train/ants_image/5650366_e22b7e1065.jpg"
img_PIL=Image.open(image_path)
img_array=np.array(img_PIL)
print(type(img_array))
print(img_array.shape)

# writer.add_image("test",img_array,1,dataformats='HWC')
# writer.add_image("test",img_array,2,dataformats='HWC')
writer.add_image("train",img_array,1,dataformats='HWC')


# for i in range(100):
#     writer.add_scalar("y=2x",3*i,i)

writer.close()

运行后来到网站查看

参考

【PyTorch深度学习快速入门教程（绝对通俗易懂！）【小土堆】】 https://www.bilibili.com/video/BV1hE411t7RN/?p=9\&share_source=copy_web\&vd_source=be33b1553b08cc7b94afdd6c8a50dc5a