如何在 PyTorch 中显示单个图像？

Posted 2023-03-12

【中文标题】如何在 PyTorch 中显示单个图像？

【英文标题】：How do I display a single image in PyTorch?

【发布时间】：2019-05-06 12:01:55

【问题描述】：

我想显示使用 ImageLoader 加载并存储在 PyTorch Tensor 中的单个图像。当我尝试通过plt.imshow(image) 显示它时，我得到：

TypeError: Invalid dimensions for image data

张量的.shape是：

torch.Size([3, 244, 244])

如何将 PyTorch 张量显示为图像？

【参考方案1】：

我编写了一个简单的函数来使用 matplotlib 可视化 pytorch 张量。

import numpy as np
import matplotlib.pyplot as plt
import torch

def show(*imgs):
    '''
     input imgs can be single or multiple tensor(s), this function uses matplotlib to visualize.
     Single input example:
     show(x) gives the visualization of x, where x should be a torch.Tensor
        if x is a 4D tensor (like image batch with the size of b(atch)*c(hannel)*h(eight)*w(eight), this function splits x in batch dimension, showing b subplots in total, where each subplot displays first 3 channels (3*h*w) at most. 
        if x is a 3D tensor, this function shows first 3 channels at most (in RGB format)
        if x is a 2D tensor, it will be shown as grayscale map
     
     Multiple input example:      
     show(x,y,z) produces three windows, displaying x, y, z respectively, where x,y,z can be in any form described above.
    '''
    img_idx = 0
    for img in imgs:
        img_idx +=1
        plt.figure(img_idx)
        if isinstance(img, torch.Tensor):
            img = img.detach().cpu()

            if img.dim()==4: # 4D tensor
                bz = img.shape[0]
                c = img.shape[1]
                if bz==1 and c==1:  # single grayscale image
                    img=img.squeeze()
                elif bz==1 and c==3: # single RGB image
                    img=img.squeeze()
                    img=img.permute(1,2,0)
                elif bz==1 and c > 3: # multiple feature maps
                    img = img[:,0:3,:,:]
                    img = img.permute(0, 2, 3, 1)[:]
                    print('warning: more than 3 channels! only channels 0,1,2 are preserved!')
                elif bz > 1 and c == 1:  # multiple grayscale images
                    img=img.squeeze()
                elif bz > 1 and c == 3:  # multiple RGB images
                    img = img.permute(0, 2, 3, 1)
                elif bz > 1 and c > 3:  # multiple feature maps
                    img = img[:,0:3,:,:]
                    img = img.permute(0, 2, 3, 1)[:]
                    print('warning: more than 3 channels! only channels 0,1,2 are preserved!')
                else:
                    raise Exception("unsupported type!  " + str(img.size()))
            elif img.dim()==3: # 3D tensor
                bz = 1
                c = img.shape[0]
                if c == 1:  # grayscale
                    img=img.squeeze()
                elif c == 3:  # RGB
                    img = img.permute(1, 2, 0)
                else:
                    raise Exception("unsupported type!  " + str(img.size()))
            elif img.dim()==2:
                pass
            else:
                raise Exception("unsupported type!  "+str(img.size()))


            img = img.numpy()  # convert to numpy
            img = img.squeeze()
            if bz ==1:
                plt.imshow(img, cmap='gray')
                # plt.colorbar()
                # plt.show()
            else:
                for idx in range(0,bz):
                    plt.subplot(int(bz**0.5),int(np.ceil(bz/int(bz**0.5))),int(idx+1))
                    plt.imshow(img[idx], cmap='gray')

        else:
            raise Exception("unsupported type:  "+str(type(img)))

【参考方案2】：

使用 fastai 的 show_image

from fastai.vision.all import show_image

【参考方案3】：

处理图像数据的 PyTorch 模块需要 C × H × W 格式的张量。¹ 而 PILlow 和 Matplotlib 期望图像数组格式为 H × W × C.²

您可以使用 TorchVision 转换轻松地将张量转换为 /from 这种格式：

from torchvision import transforms.functional as F

F.to_pil_image(image_tensor)

或者通过直接排列轴：

image_tensor.permute(1,2,0)

---

处理图像数据的 PyTorch 模块需要将张量布局为 C × H × W ：分别为通道、高度和宽度。

注意我们必须如何使用 permute 将轴的顺序从 C × H × W 更改为 H × W × C 以匹配 Matplotlib 的预期.

Deep Learning with PyTorch

【参考方案4】：

假设图像按照描述加载并存储在变量image中：

plt.imshow(transforms.ToPILImage()(image), interpolation="bicubic")
#transforms.ToPILImage()(image).show() # Alternatively

---

或者Soumith suggested:

def show(img):
    npimg = img.numpy()
    plt.imshow(np.transpose(npimg, (1, 2, 0)), interpolation='nearest')

【讨论】：

import torchvision.transforms # 可能会将导入添加到代码中

【参考方案5】：

给定图像路径名img_path的完整示例：

from PIL import Image
image = Image.open(img_path)
plt.imshow(transforms.ToPILImage()(transforms.ToTensor()(image)), interpolation="bicubic")

请注意，transforms.* 返回 class，这就是时髦的括号括起来的原因。

【参考方案6】：

给定一个代表图像的Tensor，使用.permute()将通道作为最后一个维度：

plt.imshow(  tensor_image.permute(1, 2, 0)  )

注意：permute does not copy or allocate memory 和 from_numpy() doesn't either.

【讨论】：

哇，谢谢...这对我有用...我试图做 tensor_image.numpy().reshape([224,224,3]) 并使用 cv2.imshow() 可视化它但是我没有得到实际的图像...这里出了什么问题？？

@DevashishPrasad 问题是reshape([224,224,3]) 做的事情和permute(1, 2, 0) 做的不一样。 permute 函数类似于转置矩阵，其中行变为列，列变为行。 reshape 函数做了一些完全不相关的事情，我不知道如何简洁地描述。简而言之，reshape 是错误的函数。

【参考方案7】：

正如您所见，matplotlib 即使没有转换为 numpy 数组也能正常工作。但是 PyTorch 张量（“图像张量”）是通道优先的，因此要将它们与 matplotlib 一起使用，您需要对其进行重塑：

代码：

from scipy.misc import face
import matplotlib.pyplot as plt
import torch

np_image = face()
print(type(np_image), np_image.shape)
tensor_image = torch.from_numpy(np_image)
print(type(tensor_image), tensor_image.shape)
# reshape to channel first:
tensor_image = tensor_image.view(tensor_image.shape[2], tensor_image.shape[0], tensor_image.shape[1])
print(type(tensor_image), tensor_image.shape)

# If you try to plot image with shape (C, H, W)
# You will get TypeError:
# plt.imshow(tensor_image)

# So we need to reshape it to (H, W, C):
tensor_image = tensor_image.view(tensor_image.shape[1], tensor_image.shape[2], tensor_image.shape[0])
print(type(tensor_image), tensor_image.shape)

plt.imshow(tensor_image)
plt.show()

输出：

<class 'numpy.ndarray'> (768, 1024, 3)
<class 'torch.Tensor'> torch.Size([768, 1024, 3])
<class 'torch.Tensor'> torch.Size([3, 768, 1024])
<class 'torch.Tensor'> torch.Size([768, 1024, 3])

【讨论】：

嗯，对我不起作用，请参阅张量形状的更新问题。